Oecologia DOI 10.1007/s00442-014-3173-6
PHYSIOLOGICAL ECOLOGY - ORIGINAL RESEARCH
Experimental climate warming enforces seed dormancy in South African Proteaceae but seedling drought resilience exceeds summer drought periods Judith L. Arnolds · Charles F. Musil · Anthony G. Rebelo · Gert H. J. Krüger
Received: 11 April 2014 / Accepted: 25 November 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Two hypotheses—that elevated night-time temperatures due to climate warming would enforce post-fire dormancy of Proteaceae seed due to low moisture, and that periods without rain during summer would exceed desiccation periods tolerated by Proteaceae seedlings—were tested empirically. Enforced dormancy, i.e., the inability to germinate due to an environmental restraint, was tested by measuring seed germination in 11 Proteaceae species in experimental mesocosms whose soils were artificially elevated by 1.4 and 3.5 °C above ambient by far-red wavelength filtered infrared lamps. Diminished totality of germination and velocities were observed in 91 and 64 %, respectively, of the Proteaceae species tested. Drought resilience was tested in one-year-old seedlings of 16 Proteaceae species by withholding water from potted plants during summer in a greenhouse. The most drought-resilient Proteaceae species displayed the lowest initial transpiration rates at field capacity, the smallest declines in transpiration rate with decreasing soil water content, and the lowest water losses
Communicated by Allan T. G. Green. J. L. Arnolds (*) · C. F. Musil Climate Change and Bio‑Adaptation Division, South African National Biodiversity Institute, Private Bag X7, Claremont, Cape Town 7735, South Africa e-mail: [email protected] A. G. Rebelo Applied Biodiversity Research Division, South African National Biodiversity Institute, Private Bag X7, Claremont, Cape Town 7735, South Africa G. H. J. Krüger Section Botany, School of Environmental Sciences and Development, North-West University, Potchefstroom 2520, South Africa
by transpiration. Projected drought periods leading to the complete cessation of transpiration in all Proteaceae species greatly exceeded the number of days without rain per month during summer in the current distribution ranges of those species. It was therefore concluded that enforced seed dormancy induced by elevated night-time temperatures is the post-fire recruitment stage of Proteaceae that is most sensitive to climate warming. Keywords Experimental mesocosms · Seed germination · Seedling transpiration · Soil temperature · Soil moisture
Introduction The accumulation of CO2 and other greenhouse gases in the atmosphere since pre-industrial times has already had a discernible influence on global temperature, and is predicted to cause further warming this century (IPCC 2007). On the African continent, various climate futures focusing on regional mean temperatures and rainfall changes in different seasons have been presented (Hulme et al. 2001). These draw upon different draft emission scenarios prepared for the Intergovernmental Panel on Climate Change (IPCC 2007). The scenarios predict mean annual temperature increases of between 1.3 and 4.5 °C and an up to 30 % decline in winter rainfall in the Southern African Mediterranean-climate ecosystem (Hulme et al. 2001). This ecosystem is characterized by summer drought, low soil fertility, and natural fires occurring at frequencies of 4–40 years (Kruger and Bigalke 1984). Climate-controlled differences in seedling recruitment and survival are considered the dominant influence on plant community distribution, extent, internal structure, and function (Cornelius et al. 1991). Indeed, recruitment processes
13
Oecologia
are crucial to plant species persistence, especially in the presence of regular disturbances (Klinkhamer and de Jong 1988). Although there are several processes associated with recruitment, seed germination and seedling establishment have received less attention than seed production, dispersal, and competition for limiting resources (Landhausser and Wein 1994). The Cape Floristic Region is adapted to recurrent fire cycles and characteristically experiences intense recruitment by resprouting and seed germination immediately after fire, with little or no recruitment occurring between fires (Kruger and Bigalke 1984). Fire stimulates seed release and seed germination in serotinous Proteaceae through various direct and indirect cues (Brown and Botha 2004). Serotiny is defined as an ecological adaptation in which canopy-stored seed is released in response to an environmental trigger, rather than spontaneously at seed maturation (Heelemann et al. 2008). Canopy-stored seeds of serotinous Proteaceae are dormant until activated by low temperatures (Deall and Brown 1981). Soil-stored seeds of nonserotinous Proteaceae have an additional dormancy mechanism; they can be activated by high temperatures, in order to take advantage of the early post-fire situation (Brits 1986). Increased diurnal soil temperature amplitudes resulting from the removal of the insulating vegetation layer in the immediate post-fire environment provide an important cue for some nonserotinous Proteaceae species to germinate (Brits 1987). A crucial stage following seed germination in Mediterranean-climate ecosystems is seedling establishment and persistence over the dry summer period following wintertime seedling recruitment. Germination and recruitment are critical stages in the life cycles of arid plants, with seedling survival especially vulnerable to water stress and dependent on adequate postgermination rainfall (Hoffman et al. 2009). Proteas are unique in the Cape Floristic Region in their ability to continue growing during the summer drought period, since adult plants, unlike their seedlings, are able to tap deep water. Plant sensitivity to drought is dependent upon its intensity and duration, and differs with plant species, growth stage, and organizational level (Chaves et al. 2002). Four field studies and three greenhouse studies, as summarized by Hoffman et al. (2009), have measured plant responses to drought in the Southern African Mediterranean-climate region. Field studies showed that drought reduces species richness and individual plant abundance and cover. Greenhouse studies showed that relatively short desiccation periods of 21–26 days lead to complete mortality of the evergreen shrub species P. vulgaris, L. pubescens, and G. africana. Clearly, an understanding of seed germination responses to elevated temperatures and seedling responses to drought is becoming increasingly important, as climate change scenarios suggest increases in temperature and aridity in many
13
regions of the globe (Petit et al. 1999). Also, knowledge of seed germination and seedling survival characteristics may assist in the selection of species for reintroduction into disturbed vegetation in climate mitigation programs. In view of this, we tested the following hypotheses: (1) the dormancy of Proteaceae seed is enforced by especially elevated night-time temperatures; (2) drought periods tolerated by Proteaceae seedlings are exceeded by periods without rain during summer.
Methods and materials Experiment 1: seed germination responses to experimental climate warming The experimental design comprised 15 mesocosms located outdoors in a natural setting in the Kirstenbosch National Botanical Gardens. Each mesocosm, which measured 1.2 m in length, 0.8 m in width, and 0.5 m in depth, was filled with sterile aeolian sand typical of lowland fynbos. Four infrared lamps ( PAR38 IR 175R, Philips, Amsterdam, The Netherlands) were suspended in banks above each of the 15 chambers. Far-red wavelengths emitted by the lamps were filtered with a 150-µm-thick polyethylene film (Solatrol, BP Visqueen Horticultural Products, London, UK), the average transmission of which was 38 % between 650 and 670 nm and 4 % between 720 and 740 nm (Paul et al. 2005). Silicon pyranometers interfaced with data loggers (WatchDog Series 400, Spectrum Technologies Inc., Plainfield, IL, USA) monitored radiant fluxes (range 300– 1100 nm) at the soil surface in the mesocosms at hourly intervals. Decagon 5TE sensors interfaced with Decagon E-20 loggers (Decagon Devices Inc., Pullman, WA, USA) monitored temperatures at the soil surface and volumetric moisture contents in the upper 100-mm soil layer in the mesocosms at hourly intervals. Lamps were not energized in five of the chambers; those served as ambient temperature controls. In the remaining ten chambers, which alternated with the controls, two lamps were energized in five chambers and all four lamps were energized in the remaining five chambers to provide supplementary radiant fluxes of 40 and 90 W m−2. These elevated average daily soil surface temperatures by 1.4 ± 0.03 °C in the first warming treatment (Warm 1) and 3.5 ± 0.03 °C in the second warming treatment (Warm 2), respectively, with daytime increases of a similar magnitude to the night-time increases (Fig. 1). Increases in photosynthetic photon flux densities (PPFD) at the soil surface beneath the energized infrared lamps in the mesocosms, measured with a quantum sensor (LI-189, Li-Cor, Lincoln, NE, USA), were small. These ranged between 0.8 and 1.8 µmol m−2 s−1 in the first and second warming treatments, respectively.
Oecologia
Fig. 1 Average a soil temperatures and b soil moisture contents measured hourly under ambient soil temperature and average increases of 1.4 °C (Warm 1) and 3.5 °C (Warm 2) in soil temperature
The mesocosms received natural precipitation supplemented by an automated irrigation system. Reductions in soil volumetric moisture content (Fig. 1b) in the infrared lamp-energized mesocosms were small. They averaged 0.0030 cm3 H2O cm−3 (2.0 % decrease) in the first warming treatment (Warm 1) and 0.0031 cm3 H2O cm−3 (2.1 % decrease) in the second warming treatment (Warm 2). The average night-time reductions in soil volumetric moisture content were slightly smaller than the average daytime reductions in both the first warming treatment (0.0029 cm3 H2O cm−3 night compared with 0.0031 cm3 H2O cm−3 day) and the second warming treatment (0.0025 cm3 H2O cm−3 night compared with 0.0038 cm3 H2O cm−3 day). Seeds of 11 different Proteaceae species representing three genera that were widespread in mountain and/or lowland fynbos communities were supplied by the seed division of the Kirstenbosch National Botanical Gardens. The seeds comprised composite batches collected from several different natural populations in the Mediterranean-climate Cape Floristic Region of South Africa (Mucina and
Rutherford 2006). The test species comprised the partly or totally serotinous Leucadendron coniferum (L.) Meisn., L. eucalyptifolium, H. Buek ex Meisn., L. laureolum (Lam.) Fourc., L. meridianum I. Williams, L. rubrum Burm. F., L. salicifolium (Salisb.) I. Williams, L. spissifolium subsp. spissifolium (Salisb. ex Knight) I. Williams, Protea magnifica Link., P. neriifolia R. Br., and P. repens L. as well as the nonserotinous, myrmecochorous (ant-dispersed) Leucospermum conocarpodendron (L.) H. Buek subsp. conocarpodendron. Temperature, rainfall, and elevation regimes over the natural distribution ranges of the Proteaceae test species are presented in Table 1. Only seed batches of the test Proteaceae species whose absolute germination percentages exceeded 25 % were included in the study. All seeds were initially imbibed for 12 h in diluted (1:9) smoke extract solution to promote germination (Brown and Botha 2004). The smoke extract only enhances germination once primary dormancy is overcome by seasonal temperature changes (Ooi et al. 2006). The imbibed seeds of each species were sown at the beginning of winter (June) at depths of 10 mm at 30 mm intervals into one of 11 randomized rows (25 seeds per row) spaced 70 mm apart in each of the 15 chambers. The number of seedlings of each species that emerged from each chamber was recorded at two-day intervals spanning a 90-day period after seed sowing. Mean daily germination and peak germination rates were computed for each species from sequential twice-daily counts of emerging seedlings in each treatment. Mean daily germination was calculated as the average number of seedlings emerging per day over the test period, which ended soon after the attainment of complete (absolute) germination. This provided an index of total germination that expressed both relative germination vigor and the length of the test period. Peak germination expressed the velocity of germination. This was calculated as the maximum quotient obtained from the computed cumulative germination percentage divided by the number of days from the commencement of the test. The quotients characteristically decline progressively on either side of the peak day (Czabator 1962). Experiment 2: seedling responses to drought The test seedlings comprised one-year-old seedlings, with similar heights and sizes, of 16 Proteaceae species representing four genera that were widespread in mountain and/ or lowland fynbos. The seedlings were cultivated from composite seed batches collected from several different natural populations in the Mediterranean-climate Cape Floristic Region of South Africa (Mucina and Rutherford 2006) by the seed division of the Kirstenbosch National Botanical Gardens. They included the wind-dispersed Leucadendron microcephalum (Gand.) Gand. and Schinz, L. sessile R.
13
Oecologia Table 1 Ranges of temperature, rainfall, and elevation (105 × 105 km grid cell means) over the natural distributions of the Proteaceae test species used in experiment 1 (seed germination responses to soil warming) and experiment 2 (seedling responses to drought) Species
Experiment nos.
Minimum daily temperature range (°C)
Maximum daily temperature range (°C)
Rainfall range (mm a−1)
Elevation range (m)
L. coniferum
1 and 2
4–8
21–30
275–2,217
1–797
L. eucalyptifolium
1
22–33
129–1,858
2–1,817
L. laureolum
1 and 2
−1–8 2–8
18–32
223–2,298
1–1,548
L. meridianum
1
3–8
22–28
256–948
1–703
L. microcephalum
2
2–8
19–30
318–2,298
7–1,520
L. muirii
2
5–7
24–29
256–618
1–436
L. rubrum
1
18–33
110–2,877
9–1,959
L. salicifolium
2
−3–8 1–8
19–34
123–2,815
6–1,942
L. sessile
1
2–8
19–32
442–2,373
8–1,460
L. spissifolium
1
1–8
20–31
264–1,522
13–1,558
L. catherinae
2
2–7
24–31
235–1,312
187–1,360
L. conocarpodendron subsp conocarpodendron
1 and 2
6–8
22–27
485–963
8–924
L. conocarpodendron subsp viridum
2
6–8
22–27
485–963
8–924
L. oleifolium
2
2–8
18–32
219–3,152
1–1,445
L. reflexum
2
2–8
23–33
207–1,493
1–91,339
L. truncatulum Mimetes
2
3–8
20–30
342–1,703
1–5,144
M. cucullatus
2
0–8
18–32
167–2,373
1–1,616
M. fimbriifolius
2
4–8
18–28
474–1,501
3–936
P. cynaroides
2
P. magnifica
1
−2–9
P. mundii
2
P. neriifolia
1
P. punctata
2
P. repens
1
Leucadendron
Leucospermum
Protea 18–32
208–3,345
1–1,956
−1–8
19–32
167–3,228
8–2,061
−1–9
18–33
123–3,345
7–1,705
−1–8
18–32
170–3,345
2–1,729
−3–8
20–32
129–3,138
51–1,973
−2–9
18–33
94–3,228
1–2,025
Species distributions were sourced from the ACKDAT, PRECIS (Mucina and Rutherford 2006), and Protea Atlas (Rebelo 2010) databases, and corresponding environmental data are from Schulze and Maharaj (2007), Lynch and Schulze (2007), and Schulze and Horan (2007)
Br., L. laureolum (Lam.) Fourc, L. coniferum (L.) Meisn., and L. muirii E. Phillips, the serotinous Protea punctata Meisn. P. mundii Klotzsch and P. cynaroides (L.) L., and the nonserotinous, myrmecochorous Leucospermum conocarpodendron (L.) H. Buek subsp conocarpodendron, L. conocarpodendron (L.) H. Buek subsp viridum Rourke, L. reflexum H. Buek ex Meisn., L. catherinae Compton, L. truncatum (H. Buek ex Meisn.) Rourke, L. oleifolium (P.J. Bergius) R. Br., Mimetes fimbriifolius Salisb. ex Knight, and M. cucullatus (L.) R. Br. Temperature, rainfall, and elevation regimes over the natural distribution ranges of the Proteaceae test species are presented in Table 1. Six seedlings of each of the 16 Proteaceae species (with similar ages and dimensions) were planted individually into 180 wide × 200 mm deep pots containing equivalent
13
volumes of a fynbos soil medium comprising a mixture of aeolian sand, typical of lowland fynbos, loam, and organic material (3:1:1). Pots were randomized during summer in a passively ventilated greenhouse at the Kirstenbosch National Botanical Gardens, where the summer air temperatures—monitored hourly with a radiation-shielded thermocouple interfaced with a miniature Watch Dog 450 data logger (Spectrum Technologies Inc.)—averaged 31.5 ± 0.7 °C during daytime and 21.4 ± 0.3 °C during night-time. Seedlings were allowed to acclimate in the greenhouse for a period of two weeks, during which time they were irrigated with 750 ml of water twice weekly to maintain the soils in the pots at about 50 % of field capacity. At commencement of the drought studies, all fynbos soil media in the pots were flushed with 2 L of water and
Oecologia
allowed to drain for 48 h to their field capacity of 28 % of soil volume (0.28 ml H2O cm−3 of soil), after which irrigation was withheld and measurements commenced. Soil volumetric moisture contents were measured at hourly intervals throughout the study in randomly selected pots containing each of the 16 Proteaceae test species with Decagon 5TE moisture sensors interfaced with Decagon E-20 loggers (Decagon Devices Inc.). The moisture sensors were pre-calibrated against the fynbos soil medium to improve measurement accuracy to ±1–2 % using published methods (Czarnomski et al. 2005). Transpiration rates (in mmol H2O m−2 s−1) were measured twice weekly in a 2-h period before solar noon (11:00–13:00 hours) on five fully expanded leaves of each individual plant of each Proteaceae test species with a portable leaf porometer (SC-1, Decagon Devices Inc.). Measured transpiration rates at each interval were averaged for each plant, converted to ml H2O dm2 h−1, and terminated when soil moisture contents fell below 5 % of soil volume (0.05 ml H2O cm−3 of soil). Rainfall frequency and Proteaceae distributions Long-term (1963–1999) averages of the number of days with rainfall exceeding 1 mm per month were obtained from 45 meteorological stations distributed in the southwestern Cape. Monthly rainfall frequencies were converted to days without rainfall per month and summertime (December to February) monthly averages mapped with GIS, over which were superimposed recorded distributions (105 × 105 km grid cells) of the 16 Proteaceae test species sourced from the ACKDAT, PRECIS (Mucina and Rutherford 2006), and Protea Atlas (Rebelo 2010) databases. Data synthesis and statistical analyses Measured mean daily germination and peak germination rates were log transformed to correct for non-normality in proportions. An analysis of variance tested for significant effects of elevated temperatures on the measured components of germination in each Proteaceae test species. Significantly different treatment means were separated with a Duncan’s multiple range test. Least-squares linear regressions quantified the relationship between transpiration rate (EDD) and desiccation period (DD) in days (Fig. 5a) and between transpiration rate (Esw) and volumetric water content (SW) of the soil (Fig. 5b) for each Proteaceae test species. Analyses of variance tested the regressions for statistical significance. The regression equations were used to compute the following transpiration parameters for each Proteaceae test species: (1) transpiration rate at soil field capacity (EFC), (2) average transpiration rate per drought day (AEDD), (3) average
transpiration rate per ml of soil water loss (AESW), (4) total water transpired on a drought day basis (TEDD), (5) total water transpired on a soil water basis (TESW), and (6) the drought period (in days) leading to complete cessation of transpiration (DDCT). These values were derived as follows:
DDCT = intercept/slope of regression function of EDD versus drought days
(1)
AEDD (mlH2 O dm - 2 h - 1 d - 1 ) = slope of regression function of EDD versus drought days (2)
AESW
mlH2 O dm - 2 h - 1 ml - 1 soil water loss
= [Eswt1 - Eswt2 ]/[(SWt1 × 5,090)−(SWt2 × 5,090)], (3) where Eswt1 and Eswt2 are the computed transpiration rates at the commencement (t1) and termination (t2) of transpiration measurements. SWt1 and SWt2 are the measured soil volumetric water contents at the commencement (t1) and termination (t2) of transpiration measurements. 5,090 is the volume of fynbos soil medium in a pot (in cm3). From the above computed values of AEDD and AESW, the total quantities of water transpired on a drought day basis (TEDD) and on a soil water basis (TESW) were calculated for each Proteaceae test species as follows: TEDD mlH2 O dm−2 = AEDD × 8 × DDCT (4)
TESW mlH2 O dm−2 = AESW × 8 × 1,440.8,
(5)
where DDCT is the number of drought days needed for the cessation of transpiration, 8 corresponds to the presumed 8-h daily transpiration period, and 1,440.8 is the total volume of water (in ml) present in the fynbos soil medium in each pot at field capacity. TEDD and TESW were also expressed as percentages of the total volume of water present in the fynbos soil medium at field capacity. The correspondence between all of the computed transpiration parameters was tested with the Pearson correlation coefficient and tested for significance with Student’s t test.
Results Experiment 1: seed germination responses to experimental warming Significantly (P ≤ 0.05) diminished mean daily germination (Table 2) and absolute germination (Figs. 2, 3, 4) at a
13
Oecologia Table 2 Effects of experimentally enhanced soil temperatures on germination totality and velocity of 11 South African Proteaceae species Genus/species
Mean daily germination
ANOVA
Ambient Warm 1 (+1.4 °C) Warm 2 (+3.5 °C) (F2,11)
Peak germination
ANOVA
Ambient Warm 1 (+1.4 °C) Warm 2 (+3.5 °C) (F2,11)
Leucadendron 0.929a
0.344b
0.104c
L. eucalyptifolium 0.432a 0.961a L. laureolum
0.411a
0.259b
0.536b
0.200c
0.673a
0.756a
0.479b
0.084b
5.8*
0.058b
75.9*** 1.821a
0.107b
0.095b
89.3***
L. meridianum
0.872a
0.786a
0.624b
10.9**
2.000a
2.794b
2.254a
7.6*
L. rubrum L. salicifolium
0.604a
0.344b
0.227b
11.1**
0.667a
0.418b
0.306b
8.1**
0.475a
0.360b
0.396b
4.0*
0.934a
0.786a
0.748a
L. spissifolium
0.889a
0.923a
0.855a
0.3
1.207a
2.375b
1.958c
24.5***
L. conocarpoden- 0.299a dron subsp conocarpodenron
0.013b
0.007b
71.6*** 0.348a
0.022b
0.022b
151.2***
L. coniferum
186.4*** 1.814a 4.7*
391.4***
0.4
Leucospermum
Protea P. magnifica
0.792a
0.513b
0.234c
14.1**
0.990a
0.762a
0.359b
6.5*
P. neriifolia
0.513a
0.591a
0.331b
9.7**
0.917a
1.077a
0.607b
8.7**
P. repens
0.246a
0.159b
0.159b
4.3*
0.380a
0.288a
0.302a
1.6
Treatment values with dissimilar letters are significantly different with * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001
Fig. 2 Absolute germination percentages ± standard errors measured at two-day intervals in a Leucadendron coniferum, b L. eucalyptifolium, c L. laureolum, and d L. meridianum under ambient and average 1.4 °C (Warm 1) and 3.5 °C (Warm 2) increases in soil temperature
soil temperature elevation of 3.5 °C were observed in 91 % (10) of the species, the exception being L. spissifolium, but in only 64 % (7) of the species at a soil temperature elevation of 1.4 °C (Table 2). Also, significant (P ≤ 0.05) reductions in peak germination at a soil temperature elevation of
13
3.5 °C were observed in 64 % (7) of the species but in only 36 % (4) of the species at a soil temperature elevation of 1.4 °C (Table 2). The exceptions were L. meridianum, which displayed a significantly increased (P ≤ 0.05) peak germination at 1.4 °C; L. spissifolium, which exhibited significantly
Oecologia Fig. 3 Absolute germination percentages ± standard errors measured at two-day intervals in a Leucadendron rubrum, b L. salicifolium, c L. spissifoilum, and d Leucospermum conocarpodedron subsp conocarpodendron under ambient and average 1.4 °C (Warm 1) and 3.5 °C (Warm 2) increases in soil temperature
Fig. 4 Absolute germination percentages ± standard errors measured at two-day intervals in a Protea magnifica, b P. neriifolia, and c P. repens under ambient and average increases in soil temperature of 1.4 °C (Warm 1) and 3.5 °C (Warm 2)
increased (P ≤ 0.05) peak germination at both 1.4 °C and 3.5 °C; and L. salicifolium and P. repens, which exhibited nonsignificantly (P ≥ 0.05) altered peak germination at both 1.4 and 3.5 °C (Table 2). The dormancies of L. conocarpodendron
subsp conocarpodendron and L. laureolum seeds were most severely enforced by the elevated temperatures, with a >90 % reduction in their absolute germination (Figs. 2, 3), mean daily germination, and peak germination (Table 2).
13
Oecologia
Experiment 2: seedling responses to drought
rain during summer in the current distribution of that species (Fig. 6a–d).
Measured transpiration rates declined linearly (P ≤ 0.05) with increasing duration of drought and decreasing soil volumetric water content in all 16 Proteaceae test species (Table 3). Computed initial transpiration rates at soil field capacity (EFC) and average transpiration rates per drought day (AEDD) and per ml of soil water loss (AESW) were all significantly (P ≤ 0.05) positively correlated among the 16 Proteaceae species (Table 4), as were the total quantities of water transpired on a drought day basis (TEDD) and a soil water loss basis (TESW) (Fig. 5a). However, only average transpiration rates per drought day (AEDD) were significantly (P ≤ 0.05) inversely correlated with the computed drought periods in days leading to complete cessation of transpiration (DDCT), which averaged 62.8 (range 52.3– 74.4) days in Leucadendron species, 42.1 (range 36.2– 50.1) days in Leucospermum, 40.6 (range 27.0–54.1) days in Mimetes, and 52.5 (range 49.2–61.5) days in Protea. These computed drought periods for each Proteaceae species greatly exceeded the average monthly periods without
Discussion Experiment 1: seed germination responses to experimental warming The average soil temperature increases of 1.4 and 3.5 °C recorded in the experimental warming chambers were in the range of mean annual temperature increases predicted by the SRES B1-low and A2-high climate sensitivity scenarios, respectively (means of 7-GCM experiments) for the southwestern Cape in the year 2080: between 1.3 and 4.5 °C (Hulme et al. 2001). These temperature increases resulted in significantly diminished values for absolute and mean daily germination in 91 % of the Proteaceae species tested. These were unlikely to be due to the small (2.0–2.1 %), ineffectual decreases in soil volumetric water content measured in the experimental warming chambers.
Table 3 Least-squares regressions of transpiration rate (EDD) against drought days (DD) and transpiration rate (ESW) against soil water content (SW) for 16 South African species of one-year-old Proteaceae Genus/species
Transpiration versus drought days Regression function (EDD = a × DD + b)
r
2
a
b
−1.3,131 −0.8841
68.666 50.435
0.9706 0.6573
−0.5767
36.471
−0.5582
31.288
−0.4493 −1.9236
Transpiration versus soil water content ANOVA F ratio
Regression function (ESW = c × SW + d)
r2
ANOVA F ratio
c
d
F1,16 = 231.5*** F1,21 = 15.3**
256.59
7×10−6
0.9410
F1,16 = 111.7***
183.05
0.0004
0.7708
F1,21 = 26.9***
0.9240
F1,16 = 121.5***
130.19
0.0004
0.9087
F1,16 = 99.5***
0.9783
F1,17 = 180.7***
114.98
0.0003
0.6894
F1,17 = 8.9*
33.440
0.8932
F1,23 = 58.6***
132.04
−3×10−5
0.9564
F1,23 = 153.7***
76.11
0.9088
302.98
0.0044
0.8703
F1,10 = 26.8**
48.266
0.6596
F1,10 = 39.8** F1,6 = 9.7*
161.99
−2.1682
0.7195
F1,6 = 12.8*
Leucadendron L. laureolum L. microcephalum L. coniferum L. muirii L. sessile Leucospermum L. catherinae L. conocarpodendron subsp viridum
−1.3345
43.098
0.8048
F1,12 = 33.0***
202.48
0.0013
0.8324
F1,12 = 39.7***
L. conocarpodendron subsp conocarpodenron
−0.9565
42.451
0.8828
F1,14 = 45.2***
155.86
2×10−5
0.8910
F1,14 = 49.0***
L. truncatum
−0.7871
34.897
0.8045
F1,11 = 32.9***
130.05
0.0004
0.8126
F1,11 = 34.7***
−0.4673
23.419
0.8706
F1,11 = 47.9***
84.981
0.0002
0.8308
F1,11 = 44.2***
−1.7162
49.278
0.9118
0.0004
0.8918
F1,9 = 57.7***
38.044
0.9669
F1,9 = 72.4*** F1,10 = 233.7***
189.99
−0.7037
254.17
−0.0005
0.8765
F1,10 = 56.8***
−1.0341
63.571
0.8875
0.0010
0.9145
F1,21 = 64.1***
0.8848
F1,21 = 47.3*** F1,16 = 69.1***
227.38
46.816
174.02
0.0003
0.9240
F1,16 = 109.4***
−0.6231
30.631
0.9554
F1,15 = 149.8***
109.57
0.0002
0.9582
F1,15 = 160.6***
L. reflexum
L. oleifolium Mimetes M. fimbriifolius M. cucullatus Protea P. cynaroides P. mundii P. punctata
−1.1279
−0.9982
F ratios derived from analysis of variance (ANOVA) of the regressions were significant at * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001
13
51.8 36.8 32.5 37.4
0.8356
1.1492
0.3607
0.6669
L. coniferum
L. muirii
L. sessile
43.7 57.3 44.1 36.8 24.1
0.7576
0.3118
0.8545
0.9737
0.8224
L. conocarpodendron subsp viridum
L. reflexum
L. conocarpodendron subsp conocarpodenron
L. truncatum
L. oleifolium
36.0
0.7890
M. cucullatus 64.4 49.2 31.0
0.5366
0.8979
0.6507
P. cynaroides
P. mundii
P. punctata
Protea
53.8
0.9243
M. fimbriifolius
Mimetes
85.7
0.9107
L. catherinae
Leucospermum
72.6
0.6584
0.6231
0.9982
1.0341
0.7037
1.7162
0.4673
0.7871
0.9565
1.1279
1.3345
1.9236
0.4493
0.5502
0.5767
0.8841
1.3131
245.3 (17.0)
374.5 (26.0)
508.8 (35.3)
304.6 (21.1)
370.7 (25.7)
187.3 (13.0)
278.9 (19.4)
339.7 (23.6)
344.7 (23.9)
386.5 (26.8)
609.4 (42.3)
267.4 (18.6)
250.5 (17.4)
291.6 (20.2)
403.2 (28.0)
549.4 (38.1)
49.2
46.9
61.5
54.1
27.0
50.1
44.3
44.4
38.2
36.2
39.6
74.4
56.9
63.2
57.0
52.3
EFC (ml H2O dm−2 AEDD (ml H2O dm−2 TEDD (ml H2O dm−2), % DDCT h−1 day−1) total SW in parentheses (days) h−1 at SWFC)
L. microcephalum
Sclerophylly index g dry mass dm−2
L. laureolum
Leucadendron
Genus/species
0.0215
0.0342
0.0447
0.0250
0.0300
0.0167
0.0256
0.0306
0.0398
0.0318
0.0595
0.0259
0.0226
0.0256
0.0360
0.0504
AESW (ml H2O dm−2 h−1 ml−1 SW)
248.1 (17.2)
394.1 (27.4)
514.9 (35.7)
287.8 (20.0)
346.0 (24.0)
192.4 (13.4)
294.5 (20.4)
353.0 (24.5)
458.5 (31.8)
366.8 (25.5)
686.1 (47.6)
299.0 (20.8)
260.4 (18.1)
294.8 (24.5)
414.5 (28.8)
581.1 (48.3)
TESW (ml H2O dm−2), % total SW in parentheses
Table 4 Initial transpiration rate at soil field capacity (EFC), average transpiration rates per drought day (AEDD) and per ml of soil water loss (AESW), total water transpired on either a drought day (TEDD) or a soil water (TESW) basis, and predicted drought periods for complete cessation of transpiration (DDCT) in 16 South African species of one-year-old Proteaceae
Oecologia
13
Oecologia Fig. 5 Least-squares linear regressions of a transpiration rate (EDD) against drought days (DD) for M. cucullatus juveniles, b transpiration rate (ESW) against soil moisture content (SW) for M. cucullatus juveniles, c total water transpired on a drought day basis (TEDD) against total water transpired on a soil water basis (TESW) in 16 one-year-old Proteaceae species, and d average daily transpiration rate (AEDD) against number of drought days leading to cessation of transpiration (DDCT) in 16 one-yearold Proteaceae species
Fig. 6 Average number of days without rain per month during summer, as derived from long-term (1969–1990) monthly rainfall frequencies, along with superimposed distributions of the most drought-sensitive Proteaceae test species in four different genera: a
13
Leucospermum conocarpodendron subsp viridum, b Leucadendron laureolum, c Mimetes fimbrifolius, and d Protea cynaroides. DDCT is the number of drought days required for cessation of transpiration
Oecologia
Comparison of the diurnal soil temperature minima and maxima recorded in the experimental warming chambers during the first month of germination with base and ceiling soil temperatures reported for optimum Proteaceae germination (Brits 1986) clearly indicated that the measured reductions in seedling recruitment were more closely associated with increases in diurnal soil temperature minima than maxima. The recorded average diurnal soil temperature minima of 11.0 and 12.8 °C in the first and second warming treatments both exceeded the reported base soil temperatures of 7–9 °C required for optimum Proteaceae germination (Brits 1986), as well as the highest diurnal temperature minima of between 7 and 9 °C that were evident over the natural distribution ranges of all Proteaceae test species (Table 1). This contrasted with the diurnal soil temperature maxima of 18.7 and 20.4 °C recorded in the first and second warming treatments, which were mostly below the ceiling soil temperatures of between 20 and 24 °C above which Proteaceae germination is inhibited (Brits 1986). These findings supported the proposition that the dormancy of Proteaceae seed would be enforced by especially elevated diurnal temperature minima accompanying climate warming. The ecological role of the low-temperature requirement for initiating seedling recruitment in Proteaceae has been explained as a moisture stress avoidance measure in which seed dormancy is maintained during transiently favorable rainy periods in dry Mediterraneanclimate summer months (Brits and Van Niekerk 1986). The consequences of such elevated daily temperature minima with climate warming would be the enforcement of primary dormancy in dispersed Proteaceae seeds following fire, resulting in reduced seedling recruitment during favorable wet winter months. Furthermore, the decreased germination velocity (peak values) observed with experimental warming in 64 % of the Proteaceae test species suggests that climate warming could also hinder rapid en masse seed germination after fire, a proposed means by which establishing seedlings avoid inter- and intraspecific competition with rapidly regenerating fynbos (Bond and van Wilgen 1996). Movement of dispersed Proteaceae seed to more favorable climatic locations for germination—one potential self-adaptive response to climate warming—seems highly improbable. Protea Atlas database records show that distinct Proteaceae species populations with immensely different densities exist in close proximity across elevation gradients, which is indicative of constrained seed dispersal and environmental heterogeneity. These records are supported by analyses of post-fire seeding recruitment patterns around skeletons of burnt adult Proteaceae, which show relatively short dispersal distances of only up to 17 m for ant-dispersed (myrmecochorous) seeds of some nonserotinous Proteaceae (Slingsby and Bond 1985), and between 15 and 40 m for wind-dispersed seeds of some serotinous
Proteaceae (Bond 1988; Hammill et al. 1998). More recently, empirically validated mechanistic models have demonstrated that secondary dispersal of seed along the ground by wind may also contribute significant distances to the dispersal patterns of some serotinous Proteaceae seeds after fire, with projected seed dispersal distances of up to 59 km for Protea repens, although these distances can be as short as 3 m for other serotinous species such as Leucadendron salignum (Schurr et al. 2005). However, these models assumed spatially homogeneous wind profiles and obstacle patterns. In reality, seed movement over distances of several kilometers would be restrained by impervious topographical barriers such as steep slopes, boulder fields, rivers, natural vegetation, and extensive areas that have been transformed for agricultural and forestry purposes. It is worth noting that seeds of most of the Proteaceae species sampled showed appreciable germination at both 1.4 and 3.5 °C above ambient, except for L. laureolum, which displayed less than 5 % germination at 3.5 °C above ambient, and L. conocarpodendron, with less than 2 % at 3.5 °C above ambient. Consequently, there seems to be sufficient plasticity within seeds of most of the Proteaceae species sampled to accommodate adequate seedling recruitment under the warmest climatic scenario tested. The nonserotinous myrmecochorous L. conocarpodendron differs from the serotinous Proteaceae species in that its recruitment niche (a soil-stored seed bank) probably contains a fair proportion of seeds that have remained dormant through 2–3 fire cycles (i.e., 30–80 years) in the wild, and dry stored seed 200 years old has been recorded in dry stored seed (Daws et al. 2007). However, with climate warming, these dormant seeds could gradually become less adapted to suitable germination cues. Currently, seedling germination of nonserotinous species under increases of 1.4 and 3.5 °C is exceptionally low, and effectively nonexistent for L. conocarpodendron. However, a gradual increase in temperature accompanying climate warming could allow a greater proportion of the nondormant seed bank to recruit and thus effectively select for an increasing minimum temperature threshold cue. Indeed, it took less than 50 years to inadvertently eliminate the minimum temperature threshold in cultivated Leucospermum cordifolium (G. Brits, pers. comm.), so at least some species have sufficient variability to rapidly adapt to projected temperature changes. The fact that the Cape flora has a high species richness, 9,086 species in an area of 87,892 km2 (Cowling et al. 1998), suggests that most of these species have weathered the previous climatic changes during the past glacials and interglacials. This is borne out by genetic data which suggest that the modal age for Protea is 1.4 million years (Valente et al. 2010). It also fits the scenario where most of the sister species are allopatric (Rebelo 2010), suggesting that migration was not a mechanism for coping with past climatic
13
Oecologia
changes. This is feasible in the Cape Floristic Region, with its high local relief, especially for the mountain flora, where many species would merely have to “migrate” a few hundred meters horizontally to topographically track their climatic niche. Experiment 2: seedling responses to drought The most drought-resilient Proteaceae species identified in the present study within each of the four genera displayed the lowest initial transpiration rates at field capacity, the smallest declines in transpiration rate with decreasing soil water content, and the least quantities of water transpired, and vice versa. Similarly, in comparative desiccation experiments conducted on leaves of chaparral and coastal sage species in Mediterranean-climate California, chaparral species with low initial photosynthetic and transpiration rates were also found to be more resilient to drought than coastal sage species with high initial gas and moisture exchange rates (Harrison et al. 1971). As observed in the chaparral species (Harrison et al. 1971), Proteaceae species did not display a sharp initial decrease in transpiration indicative of rapid stomatal closure, but rather a slow decline in transpiration rate with increasing soil desiccation. Such stomatal control, which assists plants in maintaining a favorable water balance during periods of drought stress (Jones 1998), has been reported in “isohydric” plants such as cowpea (Bates and Hall 1981) and maize (Tardieu et al. 1993), as well as in adult baobab tree species, even during short drought events during the rainy season (Chapotin et al. 2006). Also, it has been proposed that a high sclerophylly index—an indication of the amount of vascular and sclerenchyma tissue in the leaf—may play a role in water economy. A high concentration of mechanical cells and thick cell walls give Mediterranean sclerophyllous leaves a high leaf mass per area ratio (LMA) (Cowling and Campbell 1983), with several studies demonstrating an increase in LMA during drought (Salleo and Lo Gullo 1990; Yin 2002). Higher LMAs can be produced by increasing leaf density, leaf thickness, or both; thick leaves are often linked to a higher tolerance of drought (Chandra et al. 2004), and an increase in leaf density was also reported under low soil water availability (Groom and Lamont 1997; Niinemets 2001). However, no significant correlation was found between leaf sclerophylly and rate of water loss through transpiration by Proteaceae in this study. The projected average drought period of 49.7 ± 2.9 days required for the complete cessation of transpiration in the one-year-old Proteaceae species included in this study exceeded the mean survival periods of 26 days in sandstone soils and 22 days in shale soils reported for L. pubescens and P. vulgaris seedlings, as well as 21 days for G. africana seedlings (Hoffman et al. 2009). The capacity of the
13
organically rich, finer-textured soils in which the one-yearold Proteaceae seedlings were cultivated in this study to retain large amounts of water may partly explain their high drought resilience. Indeed, lower levels of seedling emergence in four proteoid species were reported on coarsetextured sandy soils than on adjacent fine-grained clay soils (Mustart and Cowling 1993a). This was attributed to the higher moisture capacity of the fine-grained clay soils, as corroborated by subsequent laboratory experiments where differences in the survival patterns of newly emerged seedlings of two proteoid species were found in soils with different moisture capacities (Mustart and Cowling 1993b). However, under field conditions, the recharging of soil moisture via capillary action and overnight distillation processes (Francis et al. 2007) as well as the absorption of non-rainfall moisture sources such as dew, fog, and atmospheric water vapor (Agam and Berliner 2006) could also provide additional sources of water that are captured by the extensive subsurface proteoid root system. Such moisture sources were unavailable to potted plants in the greenhouse environment. Despite this, it should be noted that there are several factors that may potentially enhance seedling resilience to drought under greenhouse conditions. These include a lower maximum transpiration rate and higher leaf water potential under greenhouse rather than field conditions due to the reduced solar irradiance and consequent smaller evaporative demand in greenhouses. This gives less weight to hydraulic than chemical mechanisms of stomatal control in the greenhouse than in the field (Kramer 1988). Also, the lower unsaturated hydraulic conductivities of natural soils in the field compared to those of soil substrates in the greenhouse increases the probability of short and local water deficits in the roots, even in wet soil and at relatively low evaporative demands (Jones and Tardieu 1998). Despite differences in drought responses between plants in natural and greenhouse environments, the projected desiccation periods leading to complete cessation of transpiration in the 16 one-year-old Proteaceae species tested in this study all greatly exceeded the number of days per month without rain in summer in their current distribution ranges (Fig. 6). Also, it was recently reported that desiccation periods of 21 days, which exceeds the average monthly period without rain during summer in over 95 % of the Cape Floristic Region (Fig. 6), still allowed complete seed germination of all 25 Proteaceae species tested in the pre-radicle emergent phase (Mustart et al. 2012). Even in the more drought-sensitive post-radicle emergent phase, a 13-day desiccation period—which exceeds the average monthly period without rain during summer in over 75 % of the Cape Floristic region (Fig. 6)—only resulted in a 50 % reduction in germination in 58 % of the Proteaceae species tested (Mustart et al. 2012). These findings of the present study lead us to conclude that Proteaceae seedlings
Oecologia
are tolerant of summer desiccation periods over a large portion of the South African Cape Floristic region, and that enforced seed dormancy induced by elevated night-time temperatures is the post-fire recruitment stage among Proteaceae that is most sensitive to climate warming. Acknowledgments We wish to sincerely thank Mr. L. Powrie for extracting Proteaceae species distributions and environmental records from various databases, and Mr. S. Snyders for assisting with the construction of experimental warming mesocosms and technical support.
References Agam N, Berliner PR (2006) Dew formation and water-vapour adsorption in semi-arid environments––a review. J Arid Environ 65:572–590 Bates LM, Hall AE (1981) Stomatal closure with soil moisture depletion not associated with changes in bulk water status. Oecologia 50:62–65 Bond WJ (1988) Proteas as “tumbleseeds”: wind dispersal through the air and over soil. S Afr J Bot 54:455–460 Bond WJ, van Wilgen BW (1996) Fire and plants. Chapman and Hall, London, p 263 Brits GJ (1986) Influence of fluctuating temperatures and H2O2 treatment on germination of Leucospermum cordifolium and Serruria florida (Proteaceae) seeds. S Afr J Bot 52:286–290 Brits GJ (1987) Germination depth versus temperature requirements in naturally dispersed seeds of Leucospermum cordifolium and L. cuneiforme (Proteaceae). S Afr J Bot 53:119–124 Brits GJ, Van Niekerk MN (1986) Effects of air temperature, oxygenating treatments and low storage temperature on seasonal germination response of Leucospermum cordifolium (Proteaceae) seeds. S Afr J Bot 52:207–211 Brits GJ, Calitz FJ, Brown NAC, Manning JC (1993) Desiccation as the active principle in heat-stimulated seed germination of Leucospermum R.Br. (Proteaceae) in fynbos. New Phytol 125:397–403 Brown NAC, Botha PA (2004) Smoke seed germination studies and a guide to seed propagation of plants from the major families of the Cape Floristic Region, South Africa. S Afr J Bot 70:559–581 Chandra A, Pathak PS, Bhatt RK, Dubey A (2004) Variation in drought tolerance of different Stylosanthes accessions. Biol Plant 48:457–460 Chapotin SM, Razanameharizaka JH, Michele Holbrook N (2006) Water relations of baobab trees (Adansonia spp L.) during the rainy season: does stem water buffer daily water deficits? Plant Cell Environ 21:1021–1032 Chaves MM, Pereira JS, Maroco J, Rodriques ML, Ricardo CPP, Osorio ML, Carvatho I, Faria T, Pinheiro C (2002) How plants cope with water stress in the field photosynthesis and growth? Ann Bot 89:907–916 Cornelius JM, Kemp PR, Ludwig JA, Cunningham GL (1991) The distribution of vascular plant species and guilds in space and time along a desert gradient. J Veg Sci 2:59–72 Cowling RM, Campbell BM (1983) The definition of leaf consistence categories in the fynbos biome and their distribution along an altitudinal gradient in the South Eastern Cape. S Afr J Bot 49:87–101 Cowling RM, Rundel PW, Desmet PG, Esler KJ (1998) Extra ordinarily high regional-scale plant diversity in southern African arid lands: sub-continental and global comparisons. Divers Distrib 4:27–36
Czabator FJ (1962) Germination value: an index combining speed and completeness of pine seed germination. For Sci 8:386–395 Czarnomski NM, Moore GW, Pypker TG, Licata J, Bond BJ (2005) Precision and accuracy of three alternative soil water content instruments in two forest soils of the Pacific Northwest. Can J For Res 35:1867–1876 Daws MI, Davies J, Vaes E, van Gelder R, Pritchard HW (2007) Two-hundred-year seed survival of Leucospermum and two other woody species from the Cape Floristic region, South Africa. Seed Sci Res 17:73–79 Deall GB, Brown NAC (1981) Seed germination in Protea magnifica Link. S Afr J Sci 77:175–176 Esler KJ, Phillips N (1994) Experimental effects of water stress on semi-arid Karoo seedlings: implication for seedling survivorship. J Arid Environ 26:325–337 Francis M, Fey M, Prinsloo H, Ellis F, Mills A, Medinski T (2007) Soils of Namaqualand: compensations of aridity. J Arid Environ 70:588–603 Groom PK, Lamont BB (1997) Xerophytic implications of increased sclerophylly: interactions with water and light in Hakea psilorryncha seedlings. New Phytol 136:231–237 Hammill KA, Bradstock RA, Allaway WG (1998) Post-fire seed dispersal and species re-establishment in proteaceous heath. Aust J Bot 46:407–419 Harrison AT, Small E, Mooney HA (1971) Drought relationships and distribution of two Mediterranean-climate California shrubs. Ecology 52:869–875 Heelemann S, Proches S, Rebelo AG, Van Wilgen B, Porembski S, Cowling RM (2008) Fire season effects on the recruitment of non-sprouting serotinous Proteaceae in the eastern (bimodal rainfall) fynbos biome, South Africa. Austral Ecol 33:119–127 Hoffman MT, Carrick PJ, Gillson L, West AG (2009) Drought, climate change and vegetation response in the succulent Karoo, South Africa. S Afr J Sci 105:54–60 Hulme M, Doherty R, Ngara T, New M, Lister D (2001) African climate change: 1900-2100. Clim Res 17:145–168 IPCC (2007) Climate change 2007: the scientific basis. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds) Contribution of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge Jones HG (1998) Stomatal control of photosynthesis and transpiration. J Exp Bot 49:387–398 Jones HG, Tardieu F (1998) Modelling water relations of horticultural crops: a review. Sci Hortic 74:21–46 Klinkhamer PGL, de Jong TJ (1988) The importance of small scale disturbance for seedling establishment in Cirsium vulgare and Cynoglossum officinale. J Ecol 76:383–392 Kramer PJ (1988) Changing concepts regarding plant water relations. Plant Cell Environ 11:565–568 Kruger FJ, Bigalke RC (1984) Fire in fynbos. In: de Booysen P, Tainton NM (eds) Ecological effects of fire in South African ecosystems. Ecol Stud vol 48. Springer, Berlin, pp 67–114 Landhausser SM, Wein RW (1994) Post fire vegetation recovery and tree establishment at the Arctic tree line: climate-change-vegetation-response hypothesis. J Ecol 81:665–672 Lynch SD, Schulze (2007) Rainfall database (Section 2.2). In: Schulze RE (ed) South African atlas of climatology and agrohydrology. WRC report 1489/1/06. Water Research Commission, Pretoria Mucina L, Rutherford MC (2006) The vegetation of South Africa, Lesotho and Swaziland. Strelitzia 19. South African National Biodiversity Institute, Pretoria Mustart PJ, Cowling RM (1993a) The role of regeneration stages in the distribution of edaphically restricted fynbos Proteaceae. Ecology 74:1490–1499
13
Oecologia Mustart PJ, Cowling RM (1993b) Effects of soil and seed characteristics on seed germination and their possible roles in determining field emergence patterns of four Agulhas Plain (South Africa) Proteaceae. Can J Bot 71:1363–1368 Mustart PJ, Rebelo AG, Juritz J, Cowling RM (2012) Wide variation in post-emergence desiccation tolerance of seedlings of fynbos proteoid shrubs. S Afr J Bot 80:110–117 Niinemets Ü (2001) Global-Scale climatic controls of leaf dry mass per area, density, and thickness in trees and shrubs. Ecology 82:453–469 Ooi MK, Auld TD, Whelan RJ (2006) Dormancy and the fire-centric focus. Germination of three Leucopogon species (Ericaceae) from south-eastern Australia. Ann Bot 98:421–430 Paul ND, Jacobson NJ, Taylor A, Wargent JJ, Moore JP (2005) The use of wavelength-selective plastic cladding materials in horticulture: understanding of crop and fungal responses through the assessment of biological spectral weighting functions. Photochem Photobiol 81:1052–1060 Petit JR, Juozel J, Raynaud D, Barkov NI, Barnola J-M, Basile I, Bender M, Chappellaz J, Davis M, Delaygue G, Delmotte M, Kotalyakov VM, Legrand M, Lipenkov VY, Lorius C, Pépin L, Ritz C, Saltzman E, Stievenard M (1999) Climate and atmospheric history from the past 420,000 years from the Vostok ice core, Antarctica. Nature 399:429–436 Rebelo T (2010) Sasol proteas––a field guide to the proteas of Southern Africa, revised edn. Fernwood, Vlaeberg Salleo S, Lo Gullo MA (1990) Sclerophylly and plant water relations in three Mediterranean Quercus species. Ann Bot 65:259–270
13
Schulze RE, Horan MJC (2007) Altitude and relative relief (Section 3.1). In: Schulze RE (ed) South African atlas of climatology and agrohydrology. WRC report 1489/1/06. Water Research Commission, Pretoria Schulze RE, Maharaj M (2007) Temperature database (Section 2.1). In: Schulze RE (ed) South African atlas of climatology and agrohydrology. WRC report 1489/1/06. Water Research Commission, Pretoria Schurr FM, Bond WJ, Midgley GF, Higgins SI (2005) A mechanistic model for secondary seed dispersal by wind and its experimental validation. J Ecol 93:1017–1028 Slingsby P, Bond WJ (1985) The influence of ants on the dispersal distance and recruitment of Leucospermum conocarpodendron (L.) Buek (Proteaceae). S Afr J Bot 51:30–34 Tardieu F, Zhang J, Gowing DJC (1993) Stomatal control by both [ABA] in the xylem sap and leaf water status: test of a model and alternative hypotheses for droughted or ABA-fed field-grown maize. Plant Cell Environ 16:413–420 Valente L, Reeves G, Schnitzler J, Mason IP, Fay MF, Rebelo TG, Chase MW, Barraclough TG (2010) Diversification of the African genus Protea (Proteaceae) in the Cape biodiversity hotspot and beyond: equal rates in different biomes. Evolution 64:745–760 Yin X (2002) Responses of leaf nitrogen concentration and specific leaf area to atmospheric CO2 enrichment: a retrospective synthesis across 62 species. Glob Change Biol 8:631–642
PHYSIOLOGICAL ECOLOGY - ORIGINAL RESEARCH
Experimental climate warming enforces seed dormancy in South African Proteaceae but seedling drought resilience exceeds summer drought periods Judith L. Arnolds · Charles F. Musil · Anthony G. Rebelo · Gert H. J. Krüger
Received: 11 April 2014 / Accepted: 25 November 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Two hypotheses—that elevated night-time temperatures due to climate warming would enforce post-fire dormancy of Proteaceae seed due to low moisture, and that periods without rain during summer would exceed desiccation periods tolerated by Proteaceae seedlings—were tested empirically. Enforced dormancy, i.e., the inability to germinate due to an environmental restraint, was tested by measuring seed germination in 11 Proteaceae species in experimental mesocosms whose soils were artificially elevated by 1.4 and 3.5 °C above ambient by far-red wavelength filtered infrared lamps. Diminished totality of germination and velocities were observed in 91 and 64 %, respectively, of the Proteaceae species tested. Drought resilience was tested in one-year-old seedlings of 16 Proteaceae species by withholding water from potted plants during summer in a greenhouse. The most drought-resilient Proteaceae species displayed the lowest initial transpiration rates at field capacity, the smallest declines in transpiration rate with decreasing soil water content, and the lowest water losses
Communicated by Allan T. G. Green. J. L. Arnolds (*) · C. F. Musil Climate Change and Bio‑Adaptation Division, South African National Biodiversity Institute, Private Bag X7, Claremont, Cape Town 7735, South Africa e-mail: [email protected] A. G. Rebelo Applied Biodiversity Research Division, South African National Biodiversity Institute, Private Bag X7, Claremont, Cape Town 7735, South Africa G. H. J. Krüger Section Botany, School of Environmental Sciences and Development, North-West University, Potchefstroom 2520, South Africa
by transpiration. Projected drought periods leading to the complete cessation of transpiration in all Proteaceae species greatly exceeded the number of days without rain per month during summer in the current distribution ranges of those species. It was therefore concluded that enforced seed dormancy induced by elevated night-time temperatures is the post-fire recruitment stage of Proteaceae that is most sensitive to climate warming. Keywords Experimental mesocosms · Seed germination · Seedling transpiration · Soil temperature · Soil moisture
Introduction The accumulation of CO2 and other greenhouse gases in the atmosphere since pre-industrial times has already had a discernible influence on global temperature, and is predicted to cause further warming this century (IPCC 2007). On the African continent, various climate futures focusing on regional mean temperatures and rainfall changes in different seasons have been presented (Hulme et al. 2001). These draw upon different draft emission scenarios prepared for the Intergovernmental Panel on Climate Change (IPCC 2007). The scenarios predict mean annual temperature increases of between 1.3 and 4.5 °C and an up to 30 % decline in winter rainfall in the Southern African Mediterranean-climate ecosystem (Hulme et al. 2001). This ecosystem is characterized by summer drought, low soil fertility, and natural fires occurring at frequencies of 4–40 years (Kruger and Bigalke 1984). Climate-controlled differences in seedling recruitment and survival are considered the dominant influence on plant community distribution, extent, internal structure, and function (Cornelius et al. 1991). Indeed, recruitment processes
13
Oecologia
are crucial to plant species persistence, especially in the presence of regular disturbances (Klinkhamer and de Jong 1988). Although there are several processes associated with recruitment, seed germination and seedling establishment have received less attention than seed production, dispersal, and competition for limiting resources (Landhausser and Wein 1994). The Cape Floristic Region is adapted to recurrent fire cycles and characteristically experiences intense recruitment by resprouting and seed germination immediately after fire, with little or no recruitment occurring between fires (Kruger and Bigalke 1984). Fire stimulates seed release and seed germination in serotinous Proteaceae through various direct and indirect cues (Brown and Botha 2004). Serotiny is defined as an ecological adaptation in which canopy-stored seed is released in response to an environmental trigger, rather than spontaneously at seed maturation (Heelemann et al. 2008). Canopy-stored seeds of serotinous Proteaceae are dormant until activated by low temperatures (Deall and Brown 1981). Soil-stored seeds of nonserotinous Proteaceae have an additional dormancy mechanism; they can be activated by high temperatures, in order to take advantage of the early post-fire situation (Brits 1986). Increased diurnal soil temperature amplitudes resulting from the removal of the insulating vegetation layer in the immediate post-fire environment provide an important cue for some nonserotinous Proteaceae species to germinate (Brits 1987). A crucial stage following seed germination in Mediterranean-climate ecosystems is seedling establishment and persistence over the dry summer period following wintertime seedling recruitment. Germination and recruitment are critical stages in the life cycles of arid plants, with seedling survival especially vulnerable to water stress and dependent on adequate postgermination rainfall (Hoffman et al. 2009). Proteas are unique in the Cape Floristic Region in their ability to continue growing during the summer drought period, since adult plants, unlike their seedlings, are able to tap deep water. Plant sensitivity to drought is dependent upon its intensity and duration, and differs with plant species, growth stage, and organizational level (Chaves et al. 2002). Four field studies and three greenhouse studies, as summarized by Hoffman et al. (2009), have measured plant responses to drought in the Southern African Mediterranean-climate region. Field studies showed that drought reduces species richness and individual plant abundance and cover. Greenhouse studies showed that relatively short desiccation periods of 21–26 days lead to complete mortality of the evergreen shrub species P. vulgaris, L. pubescens, and G. africana. Clearly, an understanding of seed germination responses to elevated temperatures and seedling responses to drought is becoming increasingly important, as climate change scenarios suggest increases in temperature and aridity in many
13
regions of the globe (Petit et al. 1999). Also, knowledge of seed germination and seedling survival characteristics may assist in the selection of species for reintroduction into disturbed vegetation in climate mitigation programs. In view of this, we tested the following hypotheses: (1) the dormancy of Proteaceae seed is enforced by especially elevated night-time temperatures; (2) drought periods tolerated by Proteaceae seedlings are exceeded by periods without rain during summer.
Methods and materials Experiment 1: seed germination responses to experimental climate warming The experimental design comprised 15 mesocosms located outdoors in a natural setting in the Kirstenbosch National Botanical Gardens. Each mesocosm, which measured 1.2 m in length, 0.8 m in width, and 0.5 m in depth, was filled with sterile aeolian sand typical of lowland fynbos. Four infrared lamps ( PAR38 IR 175R, Philips, Amsterdam, The Netherlands) were suspended in banks above each of the 15 chambers. Far-red wavelengths emitted by the lamps were filtered with a 150-µm-thick polyethylene film (Solatrol, BP Visqueen Horticultural Products, London, UK), the average transmission of which was 38 % between 650 and 670 nm and 4 % between 720 and 740 nm (Paul et al. 2005). Silicon pyranometers interfaced with data loggers (WatchDog Series 400, Spectrum Technologies Inc., Plainfield, IL, USA) monitored radiant fluxes (range 300– 1100 nm) at the soil surface in the mesocosms at hourly intervals. Decagon 5TE sensors interfaced with Decagon E-20 loggers (Decagon Devices Inc., Pullman, WA, USA) monitored temperatures at the soil surface and volumetric moisture contents in the upper 100-mm soil layer in the mesocosms at hourly intervals. Lamps were not energized in five of the chambers; those served as ambient temperature controls. In the remaining ten chambers, which alternated with the controls, two lamps were energized in five chambers and all four lamps were energized in the remaining five chambers to provide supplementary radiant fluxes of 40 and 90 W m−2. These elevated average daily soil surface temperatures by 1.4 ± 0.03 °C in the first warming treatment (Warm 1) and 3.5 ± 0.03 °C in the second warming treatment (Warm 2), respectively, with daytime increases of a similar magnitude to the night-time increases (Fig. 1). Increases in photosynthetic photon flux densities (PPFD) at the soil surface beneath the energized infrared lamps in the mesocosms, measured with a quantum sensor (LI-189, Li-Cor, Lincoln, NE, USA), were small. These ranged between 0.8 and 1.8 µmol m−2 s−1 in the first and second warming treatments, respectively.
Oecologia
Fig. 1 Average a soil temperatures and b soil moisture contents measured hourly under ambient soil temperature and average increases of 1.4 °C (Warm 1) and 3.5 °C (Warm 2) in soil temperature
The mesocosms received natural precipitation supplemented by an automated irrigation system. Reductions in soil volumetric moisture content (Fig. 1b) in the infrared lamp-energized mesocosms were small. They averaged 0.0030 cm3 H2O cm−3 (2.0 % decrease) in the first warming treatment (Warm 1) and 0.0031 cm3 H2O cm−3 (2.1 % decrease) in the second warming treatment (Warm 2). The average night-time reductions in soil volumetric moisture content were slightly smaller than the average daytime reductions in both the first warming treatment (0.0029 cm3 H2O cm−3 night compared with 0.0031 cm3 H2O cm−3 day) and the second warming treatment (0.0025 cm3 H2O cm−3 night compared with 0.0038 cm3 H2O cm−3 day). Seeds of 11 different Proteaceae species representing three genera that were widespread in mountain and/or lowland fynbos communities were supplied by the seed division of the Kirstenbosch National Botanical Gardens. The seeds comprised composite batches collected from several different natural populations in the Mediterranean-climate Cape Floristic Region of South Africa (Mucina and
Rutherford 2006). The test species comprised the partly or totally serotinous Leucadendron coniferum (L.) Meisn., L. eucalyptifolium, H. Buek ex Meisn., L. laureolum (Lam.) Fourc., L. meridianum I. Williams, L. rubrum Burm. F., L. salicifolium (Salisb.) I. Williams, L. spissifolium subsp. spissifolium (Salisb. ex Knight) I. Williams, Protea magnifica Link., P. neriifolia R. Br., and P. repens L. as well as the nonserotinous, myrmecochorous (ant-dispersed) Leucospermum conocarpodendron (L.) H. Buek subsp. conocarpodendron. Temperature, rainfall, and elevation regimes over the natural distribution ranges of the Proteaceae test species are presented in Table 1. Only seed batches of the test Proteaceae species whose absolute germination percentages exceeded 25 % were included in the study. All seeds were initially imbibed for 12 h in diluted (1:9) smoke extract solution to promote germination (Brown and Botha 2004). The smoke extract only enhances germination once primary dormancy is overcome by seasonal temperature changes (Ooi et al. 2006). The imbibed seeds of each species were sown at the beginning of winter (June) at depths of 10 mm at 30 mm intervals into one of 11 randomized rows (25 seeds per row) spaced 70 mm apart in each of the 15 chambers. The number of seedlings of each species that emerged from each chamber was recorded at two-day intervals spanning a 90-day period after seed sowing. Mean daily germination and peak germination rates were computed for each species from sequential twice-daily counts of emerging seedlings in each treatment. Mean daily germination was calculated as the average number of seedlings emerging per day over the test period, which ended soon after the attainment of complete (absolute) germination. This provided an index of total germination that expressed both relative germination vigor and the length of the test period. Peak germination expressed the velocity of germination. This was calculated as the maximum quotient obtained from the computed cumulative germination percentage divided by the number of days from the commencement of the test. The quotients characteristically decline progressively on either side of the peak day (Czabator 1962). Experiment 2: seedling responses to drought The test seedlings comprised one-year-old seedlings, with similar heights and sizes, of 16 Proteaceae species representing four genera that were widespread in mountain and/ or lowland fynbos. The seedlings were cultivated from composite seed batches collected from several different natural populations in the Mediterranean-climate Cape Floristic Region of South Africa (Mucina and Rutherford 2006) by the seed division of the Kirstenbosch National Botanical Gardens. They included the wind-dispersed Leucadendron microcephalum (Gand.) Gand. and Schinz, L. sessile R.
13
Oecologia Table 1 Ranges of temperature, rainfall, and elevation (105 × 105 km grid cell means) over the natural distributions of the Proteaceae test species used in experiment 1 (seed germination responses to soil warming) and experiment 2 (seedling responses to drought) Species
Experiment nos.
Minimum daily temperature range (°C)
Maximum daily temperature range (°C)
Rainfall range (mm a−1)
Elevation range (m)
L. coniferum
1 and 2
4–8
21–30
275–2,217
1–797
L. eucalyptifolium
1
22–33
129–1,858
2–1,817
L. laureolum
1 and 2
−1–8 2–8
18–32
223–2,298
1–1,548
L. meridianum
1
3–8
22–28
256–948
1–703
L. microcephalum
2
2–8
19–30
318–2,298
7–1,520
L. muirii
2
5–7
24–29
256–618
1–436
L. rubrum
1
18–33
110–2,877
9–1,959
L. salicifolium
2
−3–8 1–8
19–34
123–2,815
6–1,942
L. sessile
1
2–8
19–32
442–2,373
8–1,460
L. spissifolium
1
1–8
20–31
264–1,522
13–1,558
L. catherinae
2
2–7
24–31
235–1,312
187–1,360
L. conocarpodendron subsp conocarpodendron
1 and 2
6–8
22–27
485–963
8–924
L. conocarpodendron subsp viridum
2
6–8
22–27
485–963
8–924
L. oleifolium
2
2–8
18–32
219–3,152
1–1,445
L. reflexum
2
2–8
23–33
207–1,493
1–91,339
L. truncatulum Mimetes
2
3–8
20–30
342–1,703
1–5,144
M. cucullatus
2
0–8
18–32
167–2,373
1–1,616
M. fimbriifolius
2
4–8
18–28
474–1,501
3–936
P. cynaroides
2
P. magnifica
1
−2–9
P. mundii
2
P. neriifolia
1
P. punctata
2
P. repens
1
Leucadendron
Leucospermum
Protea 18–32
208–3,345
1–1,956
−1–8
19–32
167–3,228
8–2,061
−1–9
18–33
123–3,345
7–1,705
−1–8
18–32
170–3,345
2–1,729
−3–8
20–32
129–3,138
51–1,973
−2–9
18–33
94–3,228
1–2,025
Species distributions were sourced from the ACKDAT, PRECIS (Mucina and Rutherford 2006), and Protea Atlas (Rebelo 2010) databases, and corresponding environmental data are from Schulze and Maharaj (2007), Lynch and Schulze (2007), and Schulze and Horan (2007)
Br., L. laureolum (Lam.) Fourc, L. coniferum (L.) Meisn., and L. muirii E. Phillips, the serotinous Protea punctata Meisn. P. mundii Klotzsch and P. cynaroides (L.) L., and the nonserotinous, myrmecochorous Leucospermum conocarpodendron (L.) H. Buek subsp conocarpodendron, L. conocarpodendron (L.) H. Buek subsp viridum Rourke, L. reflexum H. Buek ex Meisn., L. catherinae Compton, L. truncatum (H. Buek ex Meisn.) Rourke, L. oleifolium (P.J. Bergius) R. Br., Mimetes fimbriifolius Salisb. ex Knight, and M. cucullatus (L.) R. Br. Temperature, rainfall, and elevation regimes over the natural distribution ranges of the Proteaceae test species are presented in Table 1. Six seedlings of each of the 16 Proteaceae species (with similar ages and dimensions) were planted individually into 180 wide × 200 mm deep pots containing equivalent
13
volumes of a fynbos soil medium comprising a mixture of aeolian sand, typical of lowland fynbos, loam, and organic material (3:1:1). Pots were randomized during summer in a passively ventilated greenhouse at the Kirstenbosch National Botanical Gardens, where the summer air temperatures—monitored hourly with a radiation-shielded thermocouple interfaced with a miniature Watch Dog 450 data logger (Spectrum Technologies Inc.)—averaged 31.5 ± 0.7 °C during daytime and 21.4 ± 0.3 °C during night-time. Seedlings were allowed to acclimate in the greenhouse for a period of two weeks, during which time they were irrigated with 750 ml of water twice weekly to maintain the soils in the pots at about 50 % of field capacity. At commencement of the drought studies, all fynbos soil media in the pots were flushed with 2 L of water and
Oecologia
allowed to drain for 48 h to their field capacity of 28 % of soil volume (0.28 ml H2O cm−3 of soil), after which irrigation was withheld and measurements commenced. Soil volumetric moisture contents were measured at hourly intervals throughout the study in randomly selected pots containing each of the 16 Proteaceae test species with Decagon 5TE moisture sensors interfaced with Decagon E-20 loggers (Decagon Devices Inc.). The moisture sensors were pre-calibrated against the fynbos soil medium to improve measurement accuracy to ±1–2 % using published methods (Czarnomski et al. 2005). Transpiration rates (in mmol H2O m−2 s−1) were measured twice weekly in a 2-h period before solar noon (11:00–13:00 hours) on five fully expanded leaves of each individual plant of each Proteaceae test species with a portable leaf porometer (SC-1, Decagon Devices Inc.). Measured transpiration rates at each interval were averaged for each plant, converted to ml H2O dm2 h−1, and terminated when soil moisture contents fell below 5 % of soil volume (0.05 ml H2O cm−3 of soil). Rainfall frequency and Proteaceae distributions Long-term (1963–1999) averages of the number of days with rainfall exceeding 1 mm per month were obtained from 45 meteorological stations distributed in the southwestern Cape. Monthly rainfall frequencies were converted to days without rainfall per month and summertime (December to February) monthly averages mapped with GIS, over which were superimposed recorded distributions (105 × 105 km grid cells) of the 16 Proteaceae test species sourced from the ACKDAT, PRECIS (Mucina and Rutherford 2006), and Protea Atlas (Rebelo 2010) databases. Data synthesis and statistical analyses Measured mean daily germination and peak germination rates were log transformed to correct for non-normality in proportions. An analysis of variance tested for significant effects of elevated temperatures on the measured components of germination in each Proteaceae test species. Significantly different treatment means were separated with a Duncan’s multiple range test. Least-squares linear regressions quantified the relationship between transpiration rate (EDD) and desiccation period (DD) in days (Fig. 5a) and between transpiration rate (Esw) and volumetric water content (SW) of the soil (Fig. 5b) for each Proteaceae test species. Analyses of variance tested the regressions for statistical significance. The regression equations were used to compute the following transpiration parameters for each Proteaceae test species: (1) transpiration rate at soil field capacity (EFC), (2) average transpiration rate per drought day (AEDD), (3) average
transpiration rate per ml of soil water loss (AESW), (4) total water transpired on a drought day basis (TEDD), (5) total water transpired on a soil water basis (TESW), and (6) the drought period (in days) leading to complete cessation of transpiration (DDCT). These values were derived as follows:
DDCT = intercept/slope of regression function of EDD versus drought days
(1)
AEDD (mlH2 O dm - 2 h - 1 d - 1 ) = slope of regression function of EDD versus drought days (2)
AESW
mlH2 O dm - 2 h - 1 ml - 1 soil water loss
= [Eswt1 - Eswt2 ]/[(SWt1 × 5,090)−(SWt2 × 5,090)], (3) where Eswt1 and Eswt2 are the computed transpiration rates at the commencement (t1) and termination (t2) of transpiration measurements. SWt1 and SWt2 are the measured soil volumetric water contents at the commencement (t1) and termination (t2) of transpiration measurements. 5,090 is the volume of fynbos soil medium in a pot (in cm3). From the above computed values of AEDD and AESW, the total quantities of water transpired on a drought day basis (TEDD) and on a soil water basis (TESW) were calculated for each Proteaceae test species as follows: TEDD mlH2 O dm−2 = AEDD × 8 × DDCT (4)
TESW mlH2 O dm−2 = AESW × 8 × 1,440.8,
(5)
where DDCT is the number of drought days needed for the cessation of transpiration, 8 corresponds to the presumed 8-h daily transpiration period, and 1,440.8 is the total volume of water (in ml) present in the fynbos soil medium in each pot at field capacity. TEDD and TESW were also expressed as percentages of the total volume of water present in the fynbos soil medium at field capacity. The correspondence between all of the computed transpiration parameters was tested with the Pearson correlation coefficient and tested for significance with Student’s t test.
Results Experiment 1: seed germination responses to experimental warming Significantly (P ≤ 0.05) diminished mean daily germination (Table 2) and absolute germination (Figs. 2, 3, 4) at a
13
Oecologia Table 2 Effects of experimentally enhanced soil temperatures on germination totality and velocity of 11 South African Proteaceae species Genus/species
Mean daily germination
ANOVA
Ambient Warm 1 (+1.4 °C) Warm 2 (+3.5 °C) (F2,11)
Peak germination
ANOVA
Ambient Warm 1 (+1.4 °C) Warm 2 (+3.5 °C) (F2,11)
Leucadendron 0.929a
0.344b
0.104c
L. eucalyptifolium 0.432a 0.961a L. laureolum
0.411a
0.259b
0.536b
0.200c
0.673a
0.756a
0.479b
0.084b
5.8*
0.058b
75.9*** 1.821a
0.107b
0.095b
89.3***
L. meridianum
0.872a
0.786a
0.624b
10.9**
2.000a
2.794b
2.254a
7.6*
L. rubrum L. salicifolium
0.604a
0.344b
0.227b
11.1**
0.667a
0.418b
0.306b
8.1**
0.475a
0.360b
0.396b
4.0*
0.934a
0.786a
0.748a
L. spissifolium
0.889a
0.923a
0.855a
0.3
1.207a
2.375b
1.958c
24.5***
L. conocarpoden- 0.299a dron subsp conocarpodenron
0.013b
0.007b
71.6*** 0.348a
0.022b
0.022b
151.2***
L. coniferum
186.4*** 1.814a 4.7*
391.4***
0.4
Leucospermum
Protea P. magnifica
0.792a
0.513b
0.234c
14.1**
0.990a
0.762a
0.359b
6.5*
P. neriifolia
0.513a
0.591a
0.331b
9.7**
0.917a
1.077a
0.607b
8.7**
P. repens
0.246a
0.159b
0.159b
4.3*
0.380a
0.288a
0.302a
1.6
Treatment values with dissimilar letters are significantly different with * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001
Fig. 2 Absolute germination percentages ± standard errors measured at two-day intervals in a Leucadendron coniferum, b L. eucalyptifolium, c L. laureolum, and d L. meridianum under ambient and average 1.4 °C (Warm 1) and 3.5 °C (Warm 2) increases in soil temperature
soil temperature elevation of 3.5 °C were observed in 91 % (10) of the species, the exception being L. spissifolium, but in only 64 % (7) of the species at a soil temperature elevation of 1.4 °C (Table 2). Also, significant (P ≤ 0.05) reductions in peak germination at a soil temperature elevation of
13
3.5 °C were observed in 64 % (7) of the species but in only 36 % (4) of the species at a soil temperature elevation of 1.4 °C (Table 2). The exceptions were L. meridianum, which displayed a significantly increased (P ≤ 0.05) peak germination at 1.4 °C; L. spissifolium, which exhibited significantly
Oecologia Fig. 3 Absolute germination percentages ± standard errors measured at two-day intervals in a Leucadendron rubrum, b L. salicifolium, c L. spissifoilum, and d Leucospermum conocarpodedron subsp conocarpodendron under ambient and average 1.4 °C (Warm 1) and 3.5 °C (Warm 2) increases in soil temperature
Fig. 4 Absolute germination percentages ± standard errors measured at two-day intervals in a Protea magnifica, b P. neriifolia, and c P. repens under ambient and average increases in soil temperature of 1.4 °C (Warm 1) and 3.5 °C (Warm 2)
increased (P ≤ 0.05) peak germination at both 1.4 °C and 3.5 °C; and L. salicifolium and P. repens, which exhibited nonsignificantly (P ≥ 0.05) altered peak germination at both 1.4 and 3.5 °C (Table 2). The dormancies of L. conocarpodendron
subsp conocarpodendron and L. laureolum seeds were most severely enforced by the elevated temperatures, with a >90 % reduction in their absolute germination (Figs. 2, 3), mean daily germination, and peak germination (Table 2).
13
Oecologia
Experiment 2: seedling responses to drought
rain during summer in the current distribution of that species (Fig. 6a–d).
Measured transpiration rates declined linearly (P ≤ 0.05) with increasing duration of drought and decreasing soil volumetric water content in all 16 Proteaceae test species (Table 3). Computed initial transpiration rates at soil field capacity (EFC) and average transpiration rates per drought day (AEDD) and per ml of soil water loss (AESW) were all significantly (P ≤ 0.05) positively correlated among the 16 Proteaceae species (Table 4), as were the total quantities of water transpired on a drought day basis (TEDD) and a soil water loss basis (TESW) (Fig. 5a). However, only average transpiration rates per drought day (AEDD) were significantly (P ≤ 0.05) inversely correlated with the computed drought periods in days leading to complete cessation of transpiration (DDCT), which averaged 62.8 (range 52.3– 74.4) days in Leucadendron species, 42.1 (range 36.2– 50.1) days in Leucospermum, 40.6 (range 27.0–54.1) days in Mimetes, and 52.5 (range 49.2–61.5) days in Protea. These computed drought periods for each Proteaceae species greatly exceeded the average monthly periods without
Discussion Experiment 1: seed germination responses to experimental warming The average soil temperature increases of 1.4 and 3.5 °C recorded in the experimental warming chambers were in the range of mean annual temperature increases predicted by the SRES B1-low and A2-high climate sensitivity scenarios, respectively (means of 7-GCM experiments) for the southwestern Cape in the year 2080: between 1.3 and 4.5 °C (Hulme et al. 2001). These temperature increases resulted in significantly diminished values for absolute and mean daily germination in 91 % of the Proteaceae species tested. These were unlikely to be due to the small (2.0–2.1 %), ineffectual decreases in soil volumetric water content measured in the experimental warming chambers.
Table 3 Least-squares regressions of transpiration rate (EDD) against drought days (DD) and transpiration rate (ESW) against soil water content (SW) for 16 South African species of one-year-old Proteaceae Genus/species
Transpiration versus drought days Regression function (EDD = a × DD + b)
r
2
a
b
−1.3,131 −0.8841
68.666 50.435
0.9706 0.6573
−0.5767
36.471
−0.5582
31.288
−0.4493 −1.9236
Transpiration versus soil water content ANOVA F ratio
Regression function (ESW = c × SW + d)
r2
ANOVA F ratio
c
d
F1,16 = 231.5*** F1,21 = 15.3**
256.59
7×10−6
0.9410
F1,16 = 111.7***
183.05
0.0004
0.7708
F1,21 = 26.9***
0.9240
F1,16 = 121.5***
130.19
0.0004
0.9087
F1,16 = 99.5***
0.9783
F1,17 = 180.7***
114.98
0.0003
0.6894
F1,17 = 8.9*
33.440
0.8932
F1,23 = 58.6***
132.04
−3×10−5
0.9564
F1,23 = 153.7***
76.11
0.9088
302.98
0.0044
0.8703
F1,10 = 26.8**
48.266
0.6596
F1,10 = 39.8** F1,6 = 9.7*
161.99
−2.1682
0.7195
F1,6 = 12.8*
Leucadendron L. laureolum L. microcephalum L. coniferum L. muirii L. sessile Leucospermum L. catherinae L. conocarpodendron subsp viridum
−1.3345
43.098
0.8048
F1,12 = 33.0***
202.48
0.0013
0.8324
F1,12 = 39.7***
L. conocarpodendron subsp conocarpodenron
−0.9565
42.451
0.8828
F1,14 = 45.2***
155.86
2×10−5
0.8910
F1,14 = 49.0***
L. truncatum
−0.7871
34.897
0.8045
F1,11 = 32.9***
130.05
0.0004
0.8126
F1,11 = 34.7***
−0.4673
23.419
0.8706
F1,11 = 47.9***
84.981
0.0002
0.8308
F1,11 = 44.2***
−1.7162
49.278
0.9118
0.0004
0.8918
F1,9 = 57.7***
38.044
0.9669
F1,9 = 72.4*** F1,10 = 233.7***
189.99
−0.7037
254.17
−0.0005
0.8765
F1,10 = 56.8***
−1.0341
63.571
0.8875
0.0010
0.9145
F1,21 = 64.1***
0.8848
F1,21 = 47.3*** F1,16 = 69.1***
227.38
46.816
174.02
0.0003
0.9240
F1,16 = 109.4***
−0.6231
30.631
0.9554
F1,15 = 149.8***
109.57
0.0002
0.9582
F1,15 = 160.6***
L. reflexum
L. oleifolium Mimetes M. fimbriifolius M. cucullatus Protea P. cynaroides P. mundii P. punctata
−1.1279
−0.9982
F ratios derived from analysis of variance (ANOVA) of the regressions were significant at * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001
13
51.8 36.8 32.5 37.4
0.8356
1.1492
0.3607
0.6669
L. coniferum
L. muirii
L. sessile
43.7 57.3 44.1 36.8 24.1
0.7576
0.3118
0.8545
0.9737
0.8224
L. conocarpodendron subsp viridum
L. reflexum
L. conocarpodendron subsp conocarpodenron
L. truncatum
L. oleifolium
36.0
0.7890
M. cucullatus 64.4 49.2 31.0
0.5366
0.8979
0.6507
P. cynaroides
P. mundii
P. punctata
Protea
53.8
0.9243
M. fimbriifolius
Mimetes
85.7
0.9107
L. catherinae
Leucospermum
72.6
0.6584
0.6231
0.9982
1.0341
0.7037
1.7162
0.4673
0.7871
0.9565
1.1279
1.3345
1.9236
0.4493
0.5502
0.5767
0.8841
1.3131
245.3 (17.0)
374.5 (26.0)
508.8 (35.3)
304.6 (21.1)
370.7 (25.7)
187.3 (13.0)
278.9 (19.4)
339.7 (23.6)
344.7 (23.9)
386.5 (26.8)
609.4 (42.3)
267.4 (18.6)
250.5 (17.4)
291.6 (20.2)
403.2 (28.0)
549.4 (38.1)
49.2
46.9
61.5
54.1
27.0
50.1
44.3
44.4
38.2
36.2
39.6
74.4
56.9
63.2
57.0
52.3
EFC (ml H2O dm−2 AEDD (ml H2O dm−2 TEDD (ml H2O dm−2), % DDCT h−1 day−1) total SW in parentheses (days) h−1 at SWFC)
L. microcephalum
Sclerophylly index g dry mass dm−2
L. laureolum
Leucadendron
Genus/species
0.0215
0.0342
0.0447
0.0250
0.0300
0.0167
0.0256
0.0306
0.0398
0.0318
0.0595
0.0259
0.0226
0.0256
0.0360
0.0504
AESW (ml H2O dm−2 h−1 ml−1 SW)
248.1 (17.2)
394.1 (27.4)
514.9 (35.7)
287.8 (20.0)
346.0 (24.0)
192.4 (13.4)
294.5 (20.4)
353.0 (24.5)
458.5 (31.8)
366.8 (25.5)
686.1 (47.6)
299.0 (20.8)
260.4 (18.1)
294.8 (24.5)
414.5 (28.8)
581.1 (48.3)
TESW (ml H2O dm−2), % total SW in parentheses
Table 4 Initial transpiration rate at soil field capacity (EFC), average transpiration rates per drought day (AEDD) and per ml of soil water loss (AESW), total water transpired on either a drought day (TEDD) or a soil water (TESW) basis, and predicted drought periods for complete cessation of transpiration (DDCT) in 16 South African species of one-year-old Proteaceae
Oecologia
13
Oecologia Fig. 5 Least-squares linear regressions of a transpiration rate (EDD) against drought days (DD) for M. cucullatus juveniles, b transpiration rate (ESW) against soil moisture content (SW) for M. cucullatus juveniles, c total water transpired on a drought day basis (TEDD) against total water transpired on a soil water basis (TESW) in 16 one-year-old Proteaceae species, and d average daily transpiration rate (AEDD) against number of drought days leading to cessation of transpiration (DDCT) in 16 one-yearold Proteaceae species
Fig. 6 Average number of days without rain per month during summer, as derived from long-term (1969–1990) monthly rainfall frequencies, along with superimposed distributions of the most drought-sensitive Proteaceae test species in four different genera: a
13
Leucospermum conocarpodendron subsp viridum, b Leucadendron laureolum, c Mimetes fimbrifolius, and d Protea cynaroides. DDCT is the number of drought days required for cessation of transpiration
Oecologia
Comparison of the diurnal soil temperature minima and maxima recorded in the experimental warming chambers during the first month of germination with base and ceiling soil temperatures reported for optimum Proteaceae germination (Brits 1986) clearly indicated that the measured reductions in seedling recruitment were more closely associated with increases in diurnal soil temperature minima than maxima. The recorded average diurnal soil temperature minima of 11.0 and 12.8 °C in the first and second warming treatments both exceeded the reported base soil temperatures of 7–9 °C required for optimum Proteaceae germination (Brits 1986), as well as the highest diurnal temperature minima of between 7 and 9 °C that were evident over the natural distribution ranges of all Proteaceae test species (Table 1). This contrasted with the diurnal soil temperature maxima of 18.7 and 20.4 °C recorded in the first and second warming treatments, which were mostly below the ceiling soil temperatures of between 20 and 24 °C above which Proteaceae germination is inhibited (Brits 1986). These findings supported the proposition that the dormancy of Proteaceae seed would be enforced by especially elevated diurnal temperature minima accompanying climate warming. The ecological role of the low-temperature requirement for initiating seedling recruitment in Proteaceae has been explained as a moisture stress avoidance measure in which seed dormancy is maintained during transiently favorable rainy periods in dry Mediterraneanclimate summer months (Brits and Van Niekerk 1986). The consequences of such elevated daily temperature minima with climate warming would be the enforcement of primary dormancy in dispersed Proteaceae seeds following fire, resulting in reduced seedling recruitment during favorable wet winter months. Furthermore, the decreased germination velocity (peak values) observed with experimental warming in 64 % of the Proteaceae test species suggests that climate warming could also hinder rapid en masse seed germination after fire, a proposed means by which establishing seedlings avoid inter- and intraspecific competition with rapidly regenerating fynbos (Bond and van Wilgen 1996). Movement of dispersed Proteaceae seed to more favorable climatic locations for germination—one potential self-adaptive response to climate warming—seems highly improbable. Protea Atlas database records show that distinct Proteaceae species populations with immensely different densities exist in close proximity across elevation gradients, which is indicative of constrained seed dispersal and environmental heterogeneity. These records are supported by analyses of post-fire seeding recruitment patterns around skeletons of burnt adult Proteaceae, which show relatively short dispersal distances of only up to 17 m for ant-dispersed (myrmecochorous) seeds of some nonserotinous Proteaceae (Slingsby and Bond 1985), and between 15 and 40 m for wind-dispersed seeds of some serotinous
Proteaceae (Bond 1988; Hammill et al. 1998). More recently, empirically validated mechanistic models have demonstrated that secondary dispersal of seed along the ground by wind may also contribute significant distances to the dispersal patterns of some serotinous Proteaceae seeds after fire, with projected seed dispersal distances of up to 59 km for Protea repens, although these distances can be as short as 3 m for other serotinous species such as Leucadendron salignum (Schurr et al. 2005). However, these models assumed spatially homogeneous wind profiles and obstacle patterns. In reality, seed movement over distances of several kilometers would be restrained by impervious topographical barriers such as steep slopes, boulder fields, rivers, natural vegetation, and extensive areas that have been transformed for agricultural and forestry purposes. It is worth noting that seeds of most of the Proteaceae species sampled showed appreciable germination at both 1.4 and 3.5 °C above ambient, except for L. laureolum, which displayed less than 5 % germination at 3.5 °C above ambient, and L. conocarpodendron, with less than 2 % at 3.5 °C above ambient. Consequently, there seems to be sufficient plasticity within seeds of most of the Proteaceae species sampled to accommodate adequate seedling recruitment under the warmest climatic scenario tested. The nonserotinous myrmecochorous L. conocarpodendron differs from the serotinous Proteaceae species in that its recruitment niche (a soil-stored seed bank) probably contains a fair proportion of seeds that have remained dormant through 2–3 fire cycles (i.e., 30–80 years) in the wild, and dry stored seed 200 years old has been recorded in dry stored seed (Daws et al. 2007). However, with climate warming, these dormant seeds could gradually become less adapted to suitable germination cues. Currently, seedling germination of nonserotinous species under increases of 1.4 and 3.5 °C is exceptionally low, and effectively nonexistent for L. conocarpodendron. However, a gradual increase in temperature accompanying climate warming could allow a greater proportion of the nondormant seed bank to recruit and thus effectively select for an increasing minimum temperature threshold cue. Indeed, it took less than 50 years to inadvertently eliminate the minimum temperature threshold in cultivated Leucospermum cordifolium (G. Brits, pers. comm.), so at least some species have sufficient variability to rapidly adapt to projected temperature changes. The fact that the Cape flora has a high species richness, 9,086 species in an area of 87,892 km2 (Cowling et al. 1998), suggests that most of these species have weathered the previous climatic changes during the past glacials and interglacials. This is borne out by genetic data which suggest that the modal age for Protea is 1.4 million years (Valente et al. 2010). It also fits the scenario where most of the sister species are allopatric (Rebelo 2010), suggesting that migration was not a mechanism for coping with past climatic
13
Oecologia
changes. This is feasible in the Cape Floristic Region, with its high local relief, especially for the mountain flora, where many species would merely have to “migrate” a few hundred meters horizontally to topographically track their climatic niche. Experiment 2: seedling responses to drought The most drought-resilient Proteaceae species identified in the present study within each of the four genera displayed the lowest initial transpiration rates at field capacity, the smallest declines in transpiration rate with decreasing soil water content, and the least quantities of water transpired, and vice versa. Similarly, in comparative desiccation experiments conducted on leaves of chaparral and coastal sage species in Mediterranean-climate California, chaparral species with low initial photosynthetic and transpiration rates were also found to be more resilient to drought than coastal sage species with high initial gas and moisture exchange rates (Harrison et al. 1971). As observed in the chaparral species (Harrison et al. 1971), Proteaceae species did not display a sharp initial decrease in transpiration indicative of rapid stomatal closure, but rather a slow decline in transpiration rate with increasing soil desiccation. Such stomatal control, which assists plants in maintaining a favorable water balance during periods of drought stress (Jones 1998), has been reported in “isohydric” plants such as cowpea (Bates and Hall 1981) and maize (Tardieu et al. 1993), as well as in adult baobab tree species, even during short drought events during the rainy season (Chapotin et al. 2006). Also, it has been proposed that a high sclerophylly index—an indication of the amount of vascular and sclerenchyma tissue in the leaf—may play a role in water economy. A high concentration of mechanical cells and thick cell walls give Mediterranean sclerophyllous leaves a high leaf mass per area ratio (LMA) (Cowling and Campbell 1983), with several studies demonstrating an increase in LMA during drought (Salleo and Lo Gullo 1990; Yin 2002). Higher LMAs can be produced by increasing leaf density, leaf thickness, or both; thick leaves are often linked to a higher tolerance of drought (Chandra et al. 2004), and an increase in leaf density was also reported under low soil water availability (Groom and Lamont 1997; Niinemets 2001). However, no significant correlation was found between leaf sclerophylly and rate of water loss through transpiration by Proteaceae in this study. The projected average drought period of 49.7 ± 2.9 days required for the complete cessation of transpiration in the one-year-old Proteaceae species included in this study exceeded the mean survival periods of 26 days in sandstone soils and 22 days in shale soils reported for L. pubescens and P. vulgaris seedlings, as well as 21 days for G. africana seedlings (Hoffman et al. 2009). The capacity of the
13
organically rich, finer-textured soils in which the one-yearold Proteaceae seedlings were cultivated in this study to retain large amounts of water may partly explain their high drought resilience. Indeed, lower levels of seedling emergence in four proteoid species were reported on coarsetextured sandy soils than on adjacent fine-grained clay soils (Mustart and Cowling 1993a). This was attributed to the higher moisture capacity of the fine-grained clay soils, as corroborated by subsequent laboratory experiments where differences in the survival patterns of newly emerged seedlings of two proteoid species were found in soils with different moisture capacities (Mustart and Cowling 1993b). However, under field conditions, the recharging of soil moisture via capillary action and overnight distillation processes (Francis et al. 2007) as well as the absorption of non-rainfall moisture sources such as dew, fog, and atmospheric water vapor (Agam and Berliner 2006) could also provide additional sources of water that are captured by the extensive subsurface proteoid root system. Such moisture sources were unavailable to potted plants in the greenhouse environment. Despite this, it should be noted that there are several factors that may potentially enhance seedling resilience to drought under greenhouse conditions. These include a lower maximum transpiration rate and higher leaf water potential under greenhouse rather than field conditions due to the reduced solar irradiance and consequent smaller evaporative demand in greenhouses. This gives less weight to hydraulic than chemical mechanisms of stomatal control in the greenhouse than in the field (Kramer 1988). Also, the lower unsaturated hydraulic conductivities of natural soils in the field compared to those of soil substrates in the greenhouse increases the probability of short and local water deficits in the roots, even in wet soil and at relatively low evaporative demands (Jones and Tardieu 1998). Despite differences in drought responses between plants in natural and greenhouse environments, the projected desiccation periods leading to complete cessation of transpiration in the 16 one-year-old Proteaceae species tested in this study all greatly exceeded the number of days per month without rain in summer in their current distribution ranges (Fig. 6). Also, it was recently reported that desiccation periods of 21 days, which exceeds the average monthly period without rain during summer in over 95 % of the Cape Floristic Region (Fig. 6), still allowed complete seed germination of all 25 Proteaceae species tested in the pre-radicle emergent phase (Mustart et al. 2012). Even in the more drought-sensitive post-radicle emergent phase, a 13-day desiccation period—which exceeds the average monthly period without rain during summer in over 75 % of the Cape Floristic region (Fig. 6)—only resulted in a 50 % reduction in germination in 58 % of the Proteaceae species tested (Mustart et al. 2012). These findings of the present study lead us to conclude that Proteaceae seedlings
Oecologia
are tolerant of summer desiccation periods over a large portion of the South African Cape Floristic region, and that enforced seed dormancy induced by elevated night-time temperatures is the post-fire recruitment stage among Proteaceae that is most sensitive to climate warming. Acknowledgments We wish to sincerely thank Mr. L. Powrie for extracting Proteaceae species distributions and environmental records from various databases, and Mr. S. Snyders for assisting with the construction of experimental warming mesocosms and technical support.
References Agam N, Berliner PR (2006) Dew formation and water-vapour adsorption in semi-arid environments––a review. J Arid Environ 65:572–590 Bates LM, Hall AE (1981) Stomatal closure with soil moisture depletion not associated with changes in bulk water status. Oecologia 50:62–65 Bond WJ (1988) Proteas as “tumbleseeds”: wind dispersal through the air and over soil. S Afr J Bot 54:455–460 Bond WJ, van Wilgen BW (1996) Fire and plants. Chapman and Hall, London, p 263 Brits GJ (1986) Influence of fluctuating temperatures and H2O2 treatment on germination of Leucospermum cordifolium and Serruria florida (Proteaceae) seeds. S Afr J Bot 52:286–290 Brits GJ (1987) Germination depth versus temperature requirements in naturally dispersed seeds of Leucospermum cordifolium and L. cuneiforme (Proteaceae). S Afr J Bot 53:119–124 Brits GJ, Van Niekerk MN (1986) Effects of air temperature, oxygenating treatments and low storage temperature on seasonal germination response of Leucospermum cordifolium (Proteaceae) seeds. S Afr J Bot 52:207–211 Brits GJ, Calitz FJ, Brown NAC, Manning JC (1993) Desiccation as the active principle in heat-stimulated seed germination of Leucospermum R.Br. (Proteaceae) in fynbos. New Phytol 125:397–403 Brown NAC, Botha PA (2004) Smoke seed germination studies and a guide to seed propagation of plants from the major families of the Cape Floristic Region, South Africa. S Afr J Bot 70:559–581 Chandra A, Pathak PS, Bhatt RK, Dubey A (2004) Variation in drought tolerance of different Stylosanthes accessions. Biol Plant 48:457–460 Chapotin SM, Razanameharizaka JH, Michele Holbrook N (2006) Water relations of baobab trees (Adansonia spp L.) during the rainy season: does stem water buffer daily water deficits? Plant Cell Environ 21:1021–1032 Chaves MM, Pereira JS, Maroco J, Rodriques ML, Ricardo CPP, Osorio ML, Carvatho I, Faria T, Pinheiro C (2002) How plants cope with water stress in the field photosynthesis and growth? Ann Bot 89:907–916 Cornelius JM, Kemp PR, Ludwig JA, Cunningham GL (1991) The distribution of vascular plant species and guilds in space and time along a desert gradient. J Veg Sci 2:59–72 Cowling RM, Campbell BM (1983) The definition of leaf consistence categories in the fynbos biome and their distribution along an altitudinal gradient in the South Eastern Cape. S Afr J Bot 49:87–101 Cowling RM, Rundel PW, Desmet PG, Esler KJ (1998) Extra ordinarily high regional-scale plant diversity in southern African arid lands: sub-continental and global comparisons. Divers Distrib 4:27–36
Czabator FJ (1962) Germination value: an index combining speed and completeness of pine seed germination. For Sci 8:386–395 Czarnomski NM, Moore GW, Pypker TG, Licata J, Bond BJ (2005) Precision and accuracy of three alternative soil water content instruments in two forest soils of the Pacific Northwest. Can J For Res 35:1867–1876 Daws MI, Davies J, Vaes E, van Gelder R, Pritchard HW (2007) Two-hundred-year seed survival of Leucospermum and two other woody species from the Cape Floristic region, South Africa. Seed Sci Res 17:73–79 Deall GB, Brown NAC (1981) Seed germination in Protea magnifica Link. S Afr J Sci 77:175–176 Esler KJ, Phillips N (1994) Experimental effects of water stress on semi-arid Karoo seedlings: implication for seedling survivorship. J Arid Environ 26:325–337 Francis M, Fey M, Prinsloo H, Ellis F, Mills A, Medinski T (2007) Soils of Namaqualand: compensations of aridity. J Arid Environ 70:588–603 Groom PK, Lamont BB (1997) Xerophytic implications of increased sclerophylly: interactions with water and light in Hakea psilorryncha seedlings. New Phytol 136:231–237 Hammill KA, Bradstock RA, Allaway WG (1998) Post-fire seed dispersal and species re-establishment in proteaceous heath. Aust J Bot 46:407–419 Harrison AT, Small E, Mooney HA (1971) Drought relationships and distribution of two Mediterranean-climate California shrubs. Ecology 52:869–875 Heelemann S, Proches S, Rebelo AG, Van Wilgen B, Porembski S, Cowling RM (2008) Fire season effects on the recruitment of non-sprouting serotinous Proteaceae in the eastern (bimodal rainfall) fynbos biome, South Africa. Austral Ecol 33:119–127 Hoffman MT, Carrick PJ, Gillson L, West AG (2009) Drought, climate change and vegetation response in the succulent Karoo, South Africa. S Afr J Sci 105:54–60 Hulme M, Doherty R, Ngara T, New M, Lister D (2001) African climate change: 1900-2100. Clim Res 17:145–168 IPCC (2007) Climate change 2007: the scientific basis. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds) Contribution of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge Jones HG (1998) Stomatal control of photosynthesis and transpiration. J Exp Bot 49:387–398 Jones HG, Tardieu F (1998) Modelling water relations of horticultural crops: a review. Sci Hortic 74:21–46 Klinkhamer PGL, de Jong TJ (1988) The importance of small scale disturbance for seedling establishment in Cirsium vulgare and Cynoglossum officinale. J Ecol 76:383–392 Kramer PJ (1988) Changing concepts regarding plant water relations. Plant Cell Environ 11:565–568 Kruger FJ, Bigalke RC (1984) Fire in fynbos. In: de Booysen P, Tainton NM (eds) Ecological effects of fire in South African ecosystems. Ecol Stud vol 48. Springer, Berlin, pp 67–114 Landhausser SM, Wein RW (1994) Post fire vegetation recovery and tree establishment at the Arctic tree line: climate-change-vegetation-response hypothesis. J Ecol 81:665–672 Lynch SD, Schulze (2007) Rainfall database (Section 2.2). In: Schulze RE (ed) South African atlas of climatology and agrohydrology. WRC report 1489/1/06. Water Research Commission, Pretoria Mucina L, Rutherford MC (2006) The vegetation of South Africa, Lesotho and Swaziland. Strelitzia 19. South African National Biodiversity Institute, Pretoria Mustart PJ, Cowling RM (1993a) The role of regeneration stages in the distribution of edaphically restricted fynbos Proteaceae. Ecology 74:1490–1499
13
Oecologia Mustart PJ, Cowling RM (1993b) Effects of soil and seed characteristics on seed germination and their possible roles in determining field emergence patterns of four Agulhas Plain (South Africa) Proteaceae. Can J Bot 71:1363–1368 Mustart PJ, Rebelo AG, Juritz J, Cowling RM (2012) Wide variation in post-emergence desiccation tolerance of seedlings of fynbos proteoid shrubs. S Afr J Bot 80:110–117 Niinemets Ü (2001) Global-Scale climatic controls of leaf dry mass per area, density, and thickness in trees and shrubs. Ecology 82:453–469 Ooi MK, Auld TD, Whelan RJ (2006) Dormancy and the fire-centric focus. Germination of three Leucopogon species (Ericaceae) from south-eastern Australia. Ann Bot 98:421–430 Paul ND, Jacobson NJ, Taylor A, Wargent JJ, Moore JP (2005) The use of wavelength-selective plastic cladding materials in horticulture: understanding of crop and fungal responses through the assessment of biological spectral weighting functions. Photochem Photobiol 81:1052–1060 Petit JR, Juozel J, Raynaud D, Barkov NI, Barnola J-M, Basile I, Bender M, Chappellaz J, Davis M, Delaygue G, Delmotte M, Kotalyakov VM, Legrand M, Lipenkov VY, Lorius C, Pépin L, Ritz C, Saltzman E, Stievenard M (1999) Climate and atmospheric history from the past 420,000 years from the Vostok ice core, Antarctica. Nature 399:429–436 Rebelo T (2010) Sasol proteas––a field guide to the proteas of Southern Africa, revised edn. Fernwood, Vlaeberg Salleo S, Lo Gullo MA (1990) Sclerophylly and plant water relations in three Mediterranean Quercus species. Ann Bot 65:259–270
13
Schulze RE, Horan MJC (2007) Altitude and relative relief (Section 3.1). In: Schulze RE (ed) South African atlas of climatology and agrohydrology. WRC report 1489/1/06. Water Research Commission, Pretoria Schulze RE, Maharaj M (2007) Temperature database (Section 2.1). In: Schulze RE (ed) South African atlas of climatology and agrohydrology. WRC report 1489/1/06. Water Research Commission, Pretoria Schurr FM, Bond WJ, Midgley GF, Higgins SI (2005) A mechanistic model for secondary seed dispersal by wind and its experimental validation. J Ecol 93:1017–1028 Slingsby P, Bond WJ (1985) The influence of ants on the dispersal distance and recruitment of Leucospermum conocarpodendron (L.) Buek (Proteaceae). S Afr J Bot 51:30–34 Tardieu F, Zhang J, Gowing DJC (1993) Stomatal control by both [ABA] in the xylem sap and leaf water status: test of a model and alternative hypotheses for droughted or ABA-fed field-grown maize. Plant Cell Environ 16:413–420 Valente L, Reeves G, Schnitzler J, Mason IP, Fay MF, Rebelo TG, Chase MW, Barraclough TG (2010) Diversification of the African genus Protea (Proteaceae) in the Cape biodiversity hotspot and beyond: equal rates in different biomes. Evolution 64:745–760 Yin X (2002) Responses of leaf nitrogen concentration and specific leaf area to atmospheric CO2 enrichment: a retrospective synthesis across 62 species. Glob Change Biol 8:631–642
